Re: [dpdk-dev] [PATCH] examples/kni: add dev close step when kni free
On 9/14/2021 7:02 AM, Min Hu (Connor) wrote: > From: Huisong Li > > This patch adds dev_close() step to release network adapter resources > when kni free. > > Signed-off-by: Huisong Li > Signed-off-by: Min Hu (Connor) > --- > examples/kni/main.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/examples/kni/main.c b/examples/kni/main.c > index beabb3c848..2a993a0ca4 100644 > --- a/examples/kni/main.c > +++ b/examples/kni/main.c > @@ -1031,6 +1031,7 @@ kni_free_kni(uint16_t port_id) > if (ret != 0) > RTE_LOG(ERR, APP, "Failed to stop port %d: %s\n", > port_id, rte_strerror(-ret)); > + rte_eth_dev_close(port_id); > > return 0; > } > Acked-by: Ferruh Yigit Closing the port not directly related to the 'kni free', just both done before exiting the app, so I would update the patch title to something like: examples/kni: close port before exit
Re: [dpdk-dev] [PATCH 0/2] net/mlx5: fix flow indirect action reference counting
01/09/2021 10:19, Dmitry Kozlyuk: > Dmitry Kozlyuk (2): > net/mlx5: report error on indirect CT action destroy > net/mlx5: fix flow indirect action reference counting Applied in next-net-mlx, thanks.
Re: [dpdk-dev] [PATCH] net/mlx5: fix shared RSS destruction
01/09/2021 10:07, Dmitry Kozlyuk:
> Shared RSS resources were released before checking that the shared RSS
> has no more references. If it had, the destruction was aborted, leaving
> the shared RSS in an invalid state where it could no longer be used.
> Move reference counter check before resource release.
>
> Fixes: d2046c09aa64 ("net/mlx5: support shared action for RSS")
> Cc: andr...@nvidia.com
> Cc: sta...@dpdk.org
>
> Signed-off-by: Dmitry Kozlyuk
> Acked-by: Viacheslav Ovsiienko
Applied in next-net-mlx, thanks.
Re: [dpdk-dev] [PATCH] net/bonding: fix memory leak on closing device
On 9/15/2021 9:16 AM, Min Hu (Connor) wrote:
<...>
>>> 在 2021/9/15 13:08, dapengx...@intel.com 写道:
From: Dapeng Yu
If the bond device was created by vdev mode, the kvlist was not free
after the bond device was closed.
This patch fixes it.
Fixes: 144dc4739975 ("net/bonding: fix leak on remove")
Cc: sta...@dpdk.org
Signed-off-by: Dapeng Yu
> Acked-by: Min Hu (Connor)
>
Applied to dpdk-next-net/main, thanks.
[dpdk-dev] [PATCH v5 0/3] eal: add memory pre-allocation from existing files
From: Dmitry Kozlyuk Hugepage allocation from the system takes time, resulting in slow startup or sporadic delays later. Most of the time spent in kernel is zero-filling memory for security reasons, which may be irrelevant in a controlled environment. The bottleneck is memory access speed, so for speeduup the amount of memory cleared must be reduced. We propose a new EAL option --mem-file FILE1,FILE2,... to quickly allocate dirty pages from existing files and clean it as necessary. A new malloc_perf_autotest is provided to estimate the impact. More details are explained in relevant patches. v5: rebase v4: getmntent() -> getmntent_r(), better error detection (John Levon) v3: fix hugepage mount point detection v2: fix CI failures Dmitry Kozlyuk (2): eal/linux: make hugetlbfs analysis reusable app/test: add allocator performance autotest Viacheslav Ovsiienko (1): eal: add memory pre-allocation from existing files app/test/meson.build | 2 + app/test/test_malloc_perf.c | 157 + doc/guides/linux_gsg/linux_eal_parameters.rst | 17 + lib/eal/common/eal_common_dynmem.c| 6 + lib/eal/common/eal_common_options.c | 23 ++ lib/eal/common/eal_internal_cfg.h | 4 + lib/eal/common/eal_memalloc.h | 8 +- lib/eal/common/eal_options.h | 2 + lib/eal/common/malloc_elem.c | 5 + lib/eal/common/malloc_heap.h | 8 + lib/eal/common/rte_malloc.c | 16 +- lib/eal/include/rte_memory.h | 4 +- lib/eal/linux/eal.c | 28 ++ lib/eal/linux/eal_hugepage_info.c | 158 ++--- lib/eal/linux/eal_hugepage_info.h | 39 +++ lib/eal/linux/eal_memalloc.c | 328 +- 16 files changed, 735 insertions(+), 70 deletions(-) create mode 100644 app/test/test_malloc_perf.c create mode 100644 lib/eal/linux/eal_hugepage_info.h -- 2.25.1
[dpdk-dev] [PATCH v5 1/3] eal/linux: make hugetlbfs analysis reusable
From: Dmitry Kozlyuk
get_hugepage_dir() searched for a hugetlbfs mount with a given page size
using handcraft parsing of /proc/mounts and mixing traversal logic with
selecting the needed entry. Separate code to enumerate hugetlbfs mounts
to eal_hugepage_mount_walk() taking a callback that can inspect already
parsed entries. Use mntent(3) API for parsing. This allows to reuse
enumeration logic in subsequent patches.
Signed-off-by: Dmitry Kozlyuk
Reviewed-by: Viacheslav Ovsiienko
---
lib/eal/linux/eal_hugepage_info.c | 153 +++---
lib/eal/linux/eal_hugepage_info.h | 39
2 files changed, 135 insertions(+), 57 deletions(-)
create mode 100644 lib/eal/linux/eal_hugepage_info.h
diff --git a/lib/eal/linux/eal_hugepage_info.c
b/lib/eal/linux/eal_hugepage_info.c
index d97792cade..193282e779 100644
--- a/lib/eal/linux/eal_hugepage_info.c
+++ b/lib/eal/linux/eal_hugepage_info.c
@@ -12,6 +12,7 @@
#include
#include
#include
+#include
#include
#include
#include
@@ -34,6 +35,7 @@
#include "eal_private.h"
#include "eal_internal_cfg.h"
#include "eal_hugepages.h"
+#include "eal_hugepage_info.h"
#include "eal_filesystem.h"
static const char sys_dir_path[] = "/sys/kernel/mm/hugepages";
@@ -195,73 +197,110 @@ get_default_hp_size(void)
return size;
}
-static int
-get_hugepage_dir(uint64_t hugepage_sz, char *hugedir, int len)
+int
+eal_hugepage_mount_walk(eal_hugepage_mount_walk_cb *cb, void *cb_arg)
{
- enum proc_mount_fieldnames {
- DEVICE = 0,
- MOUNTPT,
- FSTYPE,
- OPTIONS,
- _FIELDNAME_MAX
- };
- static uint64_t default_size = 0;
- const char proc_mounts[] = "/proc/mounts";
- const char hugetlbfs_str[] = "hugetlbfs";
- const size_t htlbfs_str_len = sizeof(hugetlbfs_str) - 1;
- const char pagesize_opt[] = "pagesize=";
- const size_t pagesize_opt_len = sizeof(pagesize_opt) - 1;
- const char split_tok = ' ';
- char *splitstr[_FIELDNAME_MAX];
- char buf[BUFSIZ];
- int retval = -1;
- const struct internal_config *internal_conf =
- eal_get_internal_configuration();
-
- FILE *fd = fopen(proc_mounts, "r");
- if (fd == NULL)
- rte_panic("Cannot open %s\n", proc_mounts);
+ static const char PATH[] = "/proc/mounts";
+ static const char OPTION[] = "pagesize";
+
+ static uint64_t default_size;
+
+ FILE *f = NULL;
+ struct mntent mntent;
+ char strings[PATH_MAX];
+ char *hugepage_sz_str;
+ uint64_t hugepage_sz;
+ bool stopped = false;
+ int ret = -1;
+
+ f = setmntent(PATH, "r");
+ if (f == NULL) {
+ RTE_LOG(ERR, EAL, "%s(): setmntent(%s): %s\n",
+ __func__, PATH, strerror(errno));
+ goto exit;
+ }
if (default_size == 0)
default_size = get_default_hp_size();
- while (fgets(buf, sizeof(buf), fd)){
- if (rte_strsplit(buf, sizeof(buf), splitstr, _FIELDNAME_MAX,
- split_tok) != _FIELDNAME_MAX) {
- RTE_LOG(ERR, EAL, "Error parsing %s\n", proc_mounts);
- break; /* return NULL */
+ ret = 0;
+ while (getmntent_r(f, &mntent, strings, sizeof(strings)) != NULL) {
+ if (strcmp(mntent.mnt_type, "hugetlbfs") != 0)
+ continue;
+
+ hugepage_sz_str = hasmntopt(&mntent, OPTION);
+ if (hugepage_sz_str != NULL) {
+ hugepage_sz_str += strlen(OPTION) + 1; /* +1 for '=' */
+ hugepage_sz = rte_str_to_size(hugepage_sz_str);
+ if (hugepage_sz == 0) {
+ RTE_LOG(DEBUG, EAL, "Cannot parse hugepage size
from '%s' for %s\n",
+ mntent.mnt_opts,
mntent.mnt_dir);
+ continue;
+ }
+ } else {
+ RTE_LOG(DEBUG, EAL, "Hugepage filesystem at %s without
%s option\n",
+ mntent.mnt_dir, OPTION);
+ hugepage_sz = default_size;
}
- /* we have a specified --huge-dir option, only examine that dir
*/
- if (internal_conf->hugepage_dir != NULL &&
- strcmp(splitstr[MOUNTPT],
internal_conf->hugepage_dir) != 0)
- continue;
+ if (cb(mntent.mnt_dir, hugepage_sz, cb_arg) != 0) {
+ stopped = true;
+ break;
+ }
+ }
- if (strncmp(splitstr[FSTYPE], hugetlbfs_str, htlbfs_str_len) ==
0){
- const char *pagesz_str = strstr(splitstr[OPTIONS],
pagesize_opt);
+ if (ferror(f) || (!stopped && !feof(f))) {
+
[dpdk-dev] [PATCH v5 2/3] eal: add memory pre-allocation from existing files
From: Viacheslav Ovsiienko
The primary DPDK process launch might take a long time if initially
allocated memory is large. From practice allocation of 1 TB of memory
over 1 GB hugepages on Linux takes tens of seconds. Fast restart
is highly desired for some applications and launch delay presents
a problem.
The primary delay happens in this call trace:
rte_eal_init()
rte_eal_memory_init()
rte_eal_hugepage_init()
eal_dynmem_hugepage_init()
eal_memalloc_alloc_seg_bulk()
alloc_seg()
mmap()
The largest part of the time spent in mmap() is filling the memory
with zeros. Kernel does so to prevent data leakage from a process
that was last using the page. However, in a controlled environment
it may not be the issue, while performance is. (Linux-specific
MAP_UNINITIALIZED flag allows mapping without clearing, but it is
disabled in all popular distributions for the reason above.)
It is proposed to add a new EAL option: --mem-file FILE1,FILE2,...
to map hugepages "as is" from specified FILEs in hugetlbfs.
Compared to using external memory for the task, EAL option requires
no change to application code, while allowing administrator
to control hugepage sizes and their NUMA affinity.
Limitations of the feature:
* Linux-specific (only Linux maps hugepages from files).
* Incompatible with --legacy-mem (partially replaces it).
* Incompatible with --single-file-segments
(--mem-file FILEs can contain as many segments as needed).
* Incompatible with --in-memory (logically).
A warning about possible security implications is printed
when --mem-file is used.
Until this patch DPDK allocator always cleared memory on freeing,
so that it did not have to do that on allocation, while new memory
was cleared by the kernel. When --mem-file is in use, DPDK clears memory
after allocation in rte_zmalloc() and does not clean it on freeing.
Effectively user trades fast startup for occasional allocation slowdown
whenever it is absolutely necessary. When memory is recycled, it is
cleared again, which is suboptimal par se, but saves complication
of memory management.
Signed-off-by: Viacheslav Ovsiienko
Signed-off-by: Dmitry Kozlyuk
---
doc/guides/linux_gsg/linux_eal_parameters.rst | 17 +
lib/eal/common/eal_common_dynmem.c| 6 +
lib/eal/common/eal_common_options.c | 23 ++
lib/eal/common/eal_internal_cfg.h | 4 +
lib/eal/common/eal_memalloc.h | 8 +-
lib/eal/common/eal_options.h | 2 +
lib/eal/common/malloc_elem.c | 5 +
lib/eal/common/malloc_heap.h | 8 +
lib/eal/common/rte_malloc.c | 16 +-
lib/eal/include/rte_memory.h | 4 +-
lib/eal/linux/eal.c | 28 ++
lib/eal/linux/eal_hugepage_info.c | 5 +
lib/eal/linux/eal_memalloc.c | 328 +-
13 files changed, 441 insertions(+), 13 deletions(-)
diff --git a/doc/guides/linux_gsg/linux_eal_parameters.rst
b/doc/guides/linux_gsg/linux_eal_parameters.rst
index bd3977cb3d..b465feaea8 100644
--- a/doc/guides/linux_gsg/linux_eal_parameters.rst
+++ b/doc/guides/linux_gsg/linux_eal_parameters.rst
@@ -92,6 +92,23 @@ Memory-related options
Free hugepages back to system exactly as they were originally allocated.
+* ``--mem-file ``
+
+Use memory from pre-allocated files in ``hugetlbfs`` without clearing it;
+when this memory is exhausted, switch to default dynamic allocation.
+This speeds up startup compared to ``--legacy-mem`` while also avoiding
+later delays for allocating new hugepages. One downside is slowdown
+of all zeroed memory allocations. Security warning: an application
+can access contents left by previous users of hugepages. Multiple files
+can be pre-allocated in ``hugetlbfs`` with different page sizes,
+on desired NUMA nodes, using ``mount`` options and ``numactl``:
+
+--mem-file /mnt/huge-1G/node0,/mnt/huge-1G/node1,/mnt/huge-2M/extra
+
+This option is incompatible with ``--legacy-mem``, ``--in-memory``,
+and ``--single-file-segments``. Primary and secondary processes
+must specify exactly the same list of files.
+
Other options
~
diff --git a/lib/eal/common/eal_common_dynmem.c
b/lib/eal/common/eal_common_dynmem.c
index 7c5437ddfa..abcf22f097 100644
--- a/lib/eal/common/eal_common_dynmem.c
+++ b/lib/eal/common/eal_common_dynmem.c
@@ -272,6 +272,12 @@ eal_dynmem_hugepage_init(void)
internal_conf->num_hugepage_sizes) < 0)
return -1;
+#ifdef RTE_EXEC_ENV_LINUX
+ /* pre-allocate pages from --mem-file option files */
+ if (eal_memalloc_memfile_alloc(used_hp) < 0)
+ return -1;
+#endif
+
for (hp_sz_idx = 0;
hp_sz_idx < (int)internal_conf->num_hugepage_sizes;
hp_sz_idx++) {
diff --git a/lib/eal/common/eal_common_opt
[dpdk-dev] [PATCH v5 3/3] app/test: add allocator performance autotest
From: Dmitry Kozlyuk
Memory allocator performance is crucial to applications that deal
with large amount of memory or allocate frequently. DPDK allocator
performance is affected by EAL options, API used and, at least,
allocation size. New autotest is intended to be run with different
EAL options. It measures performance with a range of sizes
for dirrerent APIs: rte_malloc, rte_zmalloc, and rte_memzone_reserve.
Work distribution between allocation and deallocation depends on EAL
options. The test prints both times and total time to ease comparison.
Memory can be filled with zeroes at different points of allocation path,
but it always takes considerable fraction of overall timing. This is why
the test measures filling speed and prints how long clearing would take
for each size as a hint.
Signed-off-by: Dmitry Kozlyuk
Reviewed-by: Viacheslav Ovsiienko
---
app/test/meson.build| 2 +
app/test/test_malloc_perf.c | 157
2 files changed, 159 insertions(+)
create mode 100644 app/test/test_malloc_perf.c
diff --git a/app/test/meson.build b/app/test/meson.build
index a7611686ad..a48dc79463 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -84,6 +84,7 @@ test_sources = files(
'test_lpm6_perf.c',
'test_lpm_perf.c',
'test_malloc.c',
+'test_malloc_perf.c',
'test_mbuf.c',
'test_member.c',
'test_member_perf.c',
@@ -281,6 +282,7 @@ fast_tests = [
perf_test_names = [
'ring_perf_autotest',
+'malloc_perf_autotest',
'mempool_perf_autotest',
'memcpy_perf_autotest',
'hash_perf_autotest',
diff --git a/app/test/test_malloc_perf.c b/app/test/test_malloc_perf.c
new file mode 100644
index 00..4435894095
--- /dev/null
+++ b/app/test/test_malloc_perf.c
@@ -0,0 +1,157 @@
+#include
+#include
+#include
+#include
+#include
+#include
+
+#include "test.h"
+
+typedef void * (alloc_t)(const char *name, size_t size, unsigned int align);
+typedef void (free_t)(void *addr);
+
+static const uint64_t KB = 1 << 10;
+static const uint64_t GB = 1 << 30;
+
+static double
+tsc_to_us(uint64_t tsc, size_t runs)
+{
+ return (double)tsc / rte_get_tsc_hz() * US_PER_S / runs;
+}
+
+static int
+test_memset_perf(double *us_per_gb)
+{
+ static const size_t RUNS = 20;
+
+ void *ptr;
+ size_t i;
+ uint64_t tsc;
+
+ puts("Performance: memset");
+
+ ptr = rte_malloc(NULL, GB, 0);
+ if (ptr == NULL) {
+ printf("rte_malloc(size=%"PRIx64") failed\n", GB);
+ return -1;
+ }
+
+ tsc = rte_rdtsc_precise();
+ for (i = 0; i < RUNS; i++)
+ memset(ptr, 0, GB);
+ tsc = rte_rdtsc_precise() - tsc;
+
+ *us_per_gb = tsc_to_us(tsc, RUNS);
+ printf("Result: %f.3 GiB/s <=> %.2f us/MiB\n",
+ US_PER_S / *us_per_gb, *us_per_gb / KB);
+
+ rte_free(ptr);
+ putchar('\n');
+ return 0;
+}
+
+static int
+test_alloc_perf(const char *name, alloc_t *alloc_fn, free_t free_fn,
+ size_t max_runs, double memset_gb_us)
+{
+ static const size_t SIZES[] = {
+ 1 << 6, 1 << 7, 1 << 10, 1 << 12, 1 << 16, 1 << 20,
+ 1 << 21, 1 << 22, 1 << 24, 1 << 30 };
+
+ size_t i, j;
+ void **ptrs;
+
+ printf("Performance: %s\n", name);
+
+ ptrs = calloc(max_runs, sizeof(ptrs[0]));
+ if (ptrs == NULL) {
+ puts("Cannot allocate memory for pointers");
+ return -1;
+ }
+
+ printf("%12s%8s%12s%12s%12s%12s\n",
+ "Size (B)", "Runs", "Alloc (us)", "Free (us)",
+ "Total (us)", "memset (us)");
+ for (i = 0; i < RTE_DIM(SIZES); i++) {
+ size_t size = SIZES[i];
+ size_t runs_done;
+ uint64_t tsc_start, tsc_alloc, tsc_free;
+ double alloc_time, free_time, memset_time;
+
+ tsc_start = rte_rdtsc_precise();
+ for (j = 0; j < max_runs; j++) {
+ ptrs[j] = alloc_fn(NULL, size, 0);
+ if (ptrs[j] == NULL)
+ break;
+ }
+ tsc_alloc = rte_rdtsc_precise() - tsc_start;
+
+ if (j == 0) {
+ printf("%12zu Interrupted: out of memory.\n", size);
+ break;
+ }
+ runs_done = j;
+
+ tsc_start = rte_rdtsc_precise();
+ for (j = 0; j < runs_done && ptrs[j] != NULL; j++)
+ free_fn(ptrs[j]);
+ tsc_free = rte_rdtsc_precise() - tsc_start;
+
+ alloc_time = tsc_to_us(tsc_alloc, runs_done);
+ free_time = tsc_to_us(tsc_free, runs_done);
+ memset_time = memset_gb_us * size / GB;
+ printf("%12zu%8zu%12.2f%12.2f%12.2f%12.2f\n",
+
Re: [dpdk-dev] [PATCH v8] ethdev: add IPv4 and L4 checksum RSS offload types
On 9/15/2021 6:47 AM, Alvin Zhang wrote: > This patch defines new RSS offload types for IPv4 and > L4(TCP/UDP/SCTP) checksum, which are required when users want > to distribute packets based on the IPv4 or L4 checksum field. > > For example "flow create 0 ingress pattern eth / ipv4 / end > actions rss types ipv4-chksum end queues end / end", this flow > causes all matching packets to be distributed to queues on > basis of IPv4 checksum. > > Signed-off-by: Alvin Zhang > Reviewed-by: Qi Zhang > Acked-by: Ajit Khaparde > Acked-by: Aman Deep Singh Acked-by: Ferruh Yigit Applied to dpdk-next-net/main, thanks.
Re: [dpdk-dev] [PATCH] Enable AddressSanitizer feature on DPDK
On Mon, Sep 20, 2021 at 9:41 PM David Christensen wrote: > >>> We do not have a ppc platform, so there is no adaptation. > >>> doc/guides/prog_guide/asan.rst has stated that we currently only > >>> support Linux x86_64. You can adapt according to the following documents, > >> the main work is to modify the base address according to the platform. > >>> Documents: > >>> https://github.com/google/sanitizers/wiki/AddressSanitizer > >>> https://github.com/llvm/llvm-project/tree/main/compiler-rt > >> > >> Understand you don't have such a platform. I looked into it and suggest > >> the > >> following change in lib/eal/common/malloc_elem.h: > >> > >> #define ASAN_SHADOW_GRAIN_SIZE 8 > >> #define ASAN_SHADOW_SCALE 3 > >> #ifdef RTE_ARCH_PPC_64 > >> #define ASAN_SHADOW_OFFSET 0x0200 #else #define > >> ASAN_SHADOW_OFFSET 0x7fff8000 #endif > >> #define ASAN_MEM_FREE_FLAG 0xfd > >> #define ASAN_MEM_REDZONE_FLAG 0xfa > >> #define ASAN_MEM_TO_SHADOW(mem) (((mem) >> > >> ASAN_SHADOW_SCALE) + > >> ASAN_SHADOW_OFFSET) > >> > >> > >> This resolves the segmentation error I receive. > >> > >> Dave > >> > > > > Great, good information for dpdk asan tool. Because we can't do many tests, > > so when this patch is merged into the main line, you can submit the ppc > > architecture patch. > > If your argument is that this is x86 only then please ensure it can't be > enabled on non-x86 platforms such as ARM and PPC. I can then easily > submit a follow-on patch to enable for PPC. > > As the patch currently stands it enables ASAN on a non-tested platform > and provides an unexpected error for some users when it can easily be > avoided. I'd advise not accepting the patch as currently presented. Please make sure only x86_64 gets this code enabled. I'll wait for a new revision, thanks. -- David Marchand
Re: [dpdk-dev] [PATCH] net/octeontx: fix invalid access to indirect buffers
On Mon, Sep 20, 2021 at 8:19 PM Harman Kalra wrote:
>
> Issue has been observed where fields of indirect buffers are
> accessed after being set free by the diver. Also fixing freeing
> of direct buffers to correct aura.
>
> Fixes: 5cbe184802aa ("net/octeontx: support fast mbuf free")
> Cc: sta...@dpdk.org
>
> Signed-off-by: David George
> Signed-off-by: Harman Kalra
Acked-by: Jerin Jacob
Applied to dpdk-next-net-mrvl/for-next-net. Thanks
> ---
> drivers/net/octeontx/octeontx_rxtx.h | 69 ++--
> 1 file changed, 46 insertions(+), 23 deletions(-)
>
> diff --git a/drivers/net/octeontx/octeontx_rxtx.h
> b/drivers/net/octeontx/octeontx_rxtx.h
> index 2ed28ea563..e0723ac26a 100644
> --- a/drivers/net/octeontx/octeontx_rxtx.h
> +++ b/drivers/net/octeontx/octeontx_rxtx.h
> @@ -161,7 +161,7 @@ ptype_table[PTYPE_SIZE][PTYPE_SIZE][PTYPE_SIZE] = {
>
>
> static __rte_always_inline uint64_t
> -octeontx_pktmbuf_detach(struct rte_mbuf *m)
> +octeontx_pktmbuf_detach(struct rte_mbuf *m, struct rte_mbuf **m_tofree)
> {
> struct rte_mempool *mp = m->pool;
> uint32_t mbuf_size, buf_len;
> @@ -171,6 +171,8 @@ octeontx_pktmbuf_detach(struct rte_mbuf *m)
>
> /* Update refcount of direct mbuf */
> md = rte_mbuf_from_indirect(m);
> + /* The real data will be in the direct buffer, inform callers this */
> + *m_tofree = md;
> refcount = rte_mbuf_refcnt_update(md, -1);
>
> priv_size = rte_pktmbuf_priv_size(mp);
> @@ -203,18 +205,18 @@ octeontx_pktmbuf_detach(struct rte_mbuf *m)
> }
>
> static __rte_always_inline uint64_t
> -octeontx_prefree_seg(struct rte_mbuf *m)
> +octeontx_prefree_seg(struct rte_mbuf *m, struct rte_mbuf **m_tofree)
> {
> if (likely(rte_mbuf_refcnt_read(m) == 1)) {
> if (!RTE_MBUF_DIRECT(m))
> - return octeontx_pktmbuf_detach(m);
> + return octeontx_pktmbuf_detach(m, m_tofree);
>
> m->next = NULL;
> m->nb_segs = 1;
> return 0;
> } else if (rte_mbuf_refcnt_update(m, -1) == 0) {
> if (!RTE_MBUF_DIRECT(m))
> - return octeontx_pktmbuf_detach(m);
> + return octeontx_pktmbuf_detach(m, m_tofree);
>
> rte_mbuf_refcnt_set(m, 1);
> m->next = NULL;
> @@ -315,6 +317,14 @@ __octeontx_xmit_prepare(struct rte_mbuf *tx_pkt,
> uint64_t *cmd_buf,
> const uint16_t flag)
> {
> uint16_t gaura_id, nb_desc = 0;
> + struct rte_mbuf *m_tofree;
> + rte_iova_t iova;
> + uint16_t data_len;
> +
> + m_tofree = tx_pkt;
> +
> + data_len = tx_pkt->data_len;
> + iova = rte_mbuf_data_iova(tx_pkt);
>
> /* Setup PKO_SEND_HDR_S */
> cmd_buf[nb_desc++] = tx_pkt->data_len & 0x;
> @@ -329,22 +339,23 @@ __octeontx_xmit_prepare(struct rte_mbuf *tx_pkt,
> uint64_t *cmd_buf,
> * not, as SG_DESC[I] and SEND_HDR[II] are clear.
> */
> if (flag & OCCTX_TX_OFFLOAD_MBUF_NOFF_F)
> - cmd_buf[0] |= (octeontx_prefree_seg(tx_pkt) <<
> + cmd_buf[0] |= (octeontx_prefree_seg(tx_pkt, &m_tofree) <<
>58);
>
> /* Mark mempool object as "put" since it is freed by PKO */
> if (!(cmd_buf[0] & (1ULL << 58)))
> - __mempool_check_cookies(tx_pkt->pool, (void **)&tx_pkt,
> + __mempool_check_cookies(m_tofree->pool, (void **)&m_tofree,
> 1, 0);
> /* Get the gaura Id */
> - gaura_id =
> octeontx_fpa_bufpool_gaura((uintptr_t)tx_pkt->pool->pool_id);
> + gaura_id =
> +
> octeontx_fpa_bufpool_gaura((uintptr_t)m_tofree->pool->pool_id);
>
> /* Setup PKO_SEND_BUFLINK_S */
> cmd_buf[nb_desc++] = PKO_SEND_BUFLINK_SUBDC |
> PKO_SEND_BUFLINK_LDTYPE(0x1ull) |
> PKO_SEND_BUFLINK_GAUAR((long)gaura_id) |
> - tx_pkt->data_len;
> - cmd_buf[nb_desc++] = rte_mbuf_data_iova(tx_pkt);
> + data_len;
> + cmd_buf[nb_desc++] = iova;
>
> return nb_desc;
> }
> @@ -355,7 +366,9 @@ __octeontx_xmit_mseg_prepare(struct rte_mbuf *tx_pkt,
> uint64_t *cmd_buf,
> {
> uint16_t nb_segs, nb_desc = 0;
> uint16_t gaura_id, len = 0;
> - struct rte_mbuf *m_next = NULL;
> + struct rte_mbuf *m_next = NULL, *m_tofree;
> + rte_iova_t iova;
> + uint16_t data_len;
>
> nb_segs = tx_pkt->nb_segs;
> /* Setup PKO_SEND_HDR_S */
> @@ -369,40 +382,50 @@ __octeontx_xmit_mseg_prepare(struct rte_mbuf *tx_pkt,
> uint64_t *cmd_buf,
>
> do {
> m_next = tx_pkt->next;
> - /* To handle case where mbufs belong to diff pools, like
> -* fragmentation
> + /* Get TX parameters up front, octeontx_prefree_seg might
> change
> +
Re: [dpdk-dev] [PATCH v4 00/13] enhancements to host based flow table management
On Mon, Sep 20, 2021 at 12:42 AM Venkat Duvvuru wrote: > > This patch set adds support for new offload features/enhancments for > Thor adapters like VF representor support, new flow matches/actions > & dynamic SRAM manager support. > > Farah Smith (4): > net/bnxt: updates to TF core index table > net/bnxt: add SRAM manager model > net/bnxt: change log level to debug > net/bnxt: add SRAM manager shared session > > Jay Ding (1): > net/bnxt: add flow meter drop counter support > > Kishore Padmanabha (6): > net/bnxt: add flow template support for Thor > net/bnxt: add support for tunnel offload API > net/bnxt: add support for dynamic encap action > net/bnxt: add wild card TCAM byte order for Thor > net/bnxt: add flow templates for Thor > net/bnxt: add enhancements to TF ULP > > Peter Spreadborough (1): > net/bnxt: enable dpool allocator > > Randy Schacher (1): > net/bnxt: dynamically allocate space for EM defrag function For the patchset, Acked-by: Ajit Khaparde Patchset applied to dpdk-next-net-brcm. 1) Changed instance of TF to TRUFLOW during merge in patch 1/13 and 13/13. Please apply to dpdk-next-net. Thanks > > doc/guides/rel_notes/release_21_11.rst| 6 + > drivers/net/bnxt/tf_core/cfa_resource_types.h | 5 +- > drivers/net/bnxt/tf_core/dpool.c |38 +- > drivers/net/bnxt/tf_core/ll.c | 3 + > drivers/net/bnxt/tf_core/ll.h |50 +- > drivers/net/bnxt/tf_core/meson.build | 2 + > drivers/net/bnxt/tf_core/tf_core.c| 169 +- > drivers/net/bnxt/tf_core/tf_core.h| 159 +- > drivers/net/bnxt/tf_core/tf_device.c |40 +- > drivers/net/bnxt/tf_core/tf_device.h | 137 +- > drivers/net/bnxt/tf_core/tf_device_p4.c |77 +- > drivers/net/bnxt/tf_core/tf_device_p4.h |50 +- > drivers/net/bnxt/tf_core/tf_device_p58.c | 112 +- > drivers/net/bnxt/tf_core/tf_device_p58.h |70 +- > drivers/net/bnxt/tf_core/tf_em.h |10 - > drivers/net/bnxt/tf_core/tf_em_common.c | 4 + > .../net/bnxt/tf_core/tf_em_hash_internal.c|34 - > drivers/net/bnxt/tf_core/tf_em_internal.c | 185 +- > drivers/net/bnxt/tf_core/tf_msg.c | 2 +- > drivers/net/bnxt/tf_core/tf_rm.c | 180 +- > drivers/net/bnxt/tf_core/tf_rm.h |62 +- > drivers/net/bnxt/tf_core/tf_session.c |56 + > drivers/net/bnxt/tf_core/tf_session.h |58 +- > drivers/net/bnxt/tf_core/tf_sram_mgr.c| 971 + > drivers/net/bnxt/tf_core/tf_sram_mgr.h| 317 + > drivers/net/bnxt/tf_core/tf_tbl.c | 259 +- > drivers/net/bnxt/tf_core/tf_tbl.h |87 +- > drivers/net/bnxt/tf_core/tf_tbl_sram.c| 747 + > drivers/net/bnxt/tf_core/tf_tbl_sram.h| 154 + > drivers/net/bnxt/tf_core/tf_tcam.c|16 +- > drivers/net/bnxt/tf_core/tf_tcam.h| 7 + > drivers/net/bnxt/tf_core/tf_tcam_shared.c |28 +- > drivers/net/bnxt/tf_core/tf_util.c|12 + > drivers/net/bnxt/tf_ulp/bnxt_tf_common.h |10 +- > drivers/net/bnxt/tf_ulp/bnxt_ulp.c|52 +- > drivers/net/bnxt/tf_ulp/bnxt_ulp.h|20 +- > drivers/net/bnxt/tf_ulp/bnxt_ulp_flow.c | 226 +- > .../bnxt/tf_ulp/generic_templates/meson.build | 3 + > .../generic_templates/ulp_template_db_act.c | 2 +- > .../generic_templates/ulp_template_db_class.c | 12109 +++- > .../generic_templates/ulp_template_db_enum.h | 618 +- > .../generic_templates/ulp_template_db_field.h | 767 +- > .../generic_templates/ulp_template_db_tbl.c | 2757 +- > .../ulp_template_db_thor_act.c| 5079 +- > .../ulp_template_db_thor_class.c | 45573 ++-- > .../ulp_template_db_wh_plus_act.c | 1700 +- > .../ulp_template_db_wh_plus_class.c | 8329 ++- > drivers/net/bnxt/tf_ulp/ulp_fc_mgr.c |48 +- > drivers/net/bnxt/tf_ulp/ulp_fc_mgr.h | 8 +- > drivers/net/bnxt/tf_ulp/ulp_flow_db.c | 678 +- > drivers/net/bnxt/tf_ulp/ulp_flow_db.h |68 +- > drivers/net/bnxt/tf_ulp/ulp_gen_tbl.c | 9 +- > drivers/net/bnxt/tf_ulp/ulp_mapper.c | 448 +- > drivers/net/bnxt/tf_ulp/ulp_mapper.h |10 +- > drivers/net/bnxt/tf_ulp/ulp_matcher.c |13 + > drivers/net/bnxt/tf_ulp/ulp_port_db.c |15 +- > drivers/net/bnxt/tf_ulp/ulp_rte_handler_tbl.c |31 + > drivers/net/bnxt/tf_ulp/ulp_rte_parser.c | 663 +- > drivers/net/bnxt/tf_ulp/ulp_rte_parser.h |12 +- > drivers/net/bnxt/tf_ulp/ulp_template_struct.h |32 +- > drivers/net/bnxt/tf_ulp/ulp_tun.c | 521 +- > drivers/net/bnxt/tf_ulp/ulp_tun.h |89 +- > drivers/net/bnxt/tf_ulp/ulp_utils.c |71 +- > drivers/net/bnxt/tf_ulp/ulp_utils.h |27 +-
Re: [dpdk-dev] [EXT] Re: [PATCH] eal: add telemetry callbacks for memory info
> -Original Message-
> From: Bruce Richardson
> Sent: Monday, September 20, 2021 9:27 PM
> To: Harman Kalra
> Cc: dev@dpdk.org; ciara.po...@intel.com; Anatoly Burakov
>
> Subject: [EXT] Re: [PATCH] eal: add telemetry callbacks for memory info
>
> External Email
>
> --
> On Wed, Sep 15, 2021 at 03:23:36PM +0530, Harman Kalra wrote:
> > Registering new telemetry callbacks to dump named (memzones) and
> > unnamed (malloc) memory information to a file provided as an argument.
> >
> > Example:
> > Connecting to /var/run/dpdk/rte/dpdk_telemetry.v2
> > {"version": "DPDK 21.08.0", "pid": 34075, "max_output_len": 16384}
> > Connected to application: "dpdk-testpmd"
> > --> /eal/malloc_dump,/tmp/malloc_dump
> > {"/eal/malloc_dump": {"Malloc elements file: ": "/tmp/malloc_dump"}}
> > -->
> > --> /eal/malloc_info,/tmp/info
> > {"/eal/malloc_info": {"Malloc stats file: ": "/tmp/info"}}
> > -->
> > -->
> > --> /eal/memzone_dump,/tmp/memzone_info
> > {"/eal/memzone_dump": {"Memzones count: ": 11, \ "Memzones info file:
> > ": "/tmp/memzone_info"}}
> >
> > Signed-off-by: Harman Kalra
> > ---
>
> For this info, why not just send the data out as telemetry data rather than
> writing files on the filesystem containing it? If the info is too large to
> dump it
> all in a single go, a shortened form could be sent via some form of list call,
> and additional calls could be used to provide more detail on specific items in
> the list.
>
> Also, this seems more a debugging operation than a telemetry one, though I
> don't have a strong objection to the info being exported as telemetry directly
> (just not via filesystem).
>
> Regards,
> /Bruce
Hi Bruce,
Thanks for reviewing the patch.
I have implemented these telemetry commands as a wrapper which uses existing
malloc/memzone debug APIs to
collect the debug information, these debug APIs are implemented in the way that
they accept a file pointer/stdout.
to get the information.
As a solution either I should make changes to these debug APIs to accept a
buffer also? Or other way could be get
the info dumped into a file, and inside telemetry command parse and convert the
info into json format and send it.
But its lot of debug information so will require multiple iterations as you
suggested. But on client (peer) side one
will have to again convert json to retrieve the info.
Just for my understanding, what drawback do you see in dumping the information
to a file? Because on peer side
It is very convenient to read the information from dumped file and use it.
Thanks
Harman
Re: [dpdk-dev] [PATCH v2] common/cnxk: align NPA stack to ROC cache line size
On Fri, Sep 17, 2021 at 4:55 PM Ashwin Sekhar T K wrote:
>
> Network Pool accelerator (NPA) is part of ROC (Rest Of Chip). So
> NPA structures should be aligned to ROC Cache line size and not
> CPU cache line size.
>
> Non alignment of NPA stack to ROC cache line will result in
> undefined runtime NPA behaviour.
>
> Fixes: f765f5611240 ("common/cnxk: add NPA pool HW operations")
>
> Signed-off-by: Ashwin Sekhar T K
> Acked-by: Jerin Jacob
Applied to dpdk-next-net-mrvl/for-next-net. Thanks
> ---
> drivers/common/cnxk/roc_npa.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/common/cnxk/roc_npa.c b/drivers/common/cnxk/roc_npa.c
> index d064d125c1..a0d2cc8f19 100644
> --- a/drivers/common/cnxk/roc_npa.c
> +++ b/drivers/common/cnxk/roc_npa.c
> @@ -194,7 +194,7 @@ npa_stack_dma_alloc(struct npa_lf *lf, char *name, int
> pool_id, size_t size)
> {
> const char *mz_name = npa_stack_memzone_name(lf, pool_id, name);
>
> - return plt_memzone_reserve_cache_align(mz_name, size);
> + return plt_memzone_reserve_aligned(mz_name, size, 0, ROC_ALIGN);
> }
>
> static inline int
> --
> 2.32.0
>
Re: [dpdk-dev] [PATCH] net/ice: add ability to reduce the Rx latency
On 18/09/2021 02:33, Zhang, AlvinX wrote:
-Original Message-
From: Kevin Traynor
Sent: Saturday, September 18, 2021 1:25 AM
To: Zhang, AlvinX ; Zhang, Qi Z
; Guo, Junfeng
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH] net/ice: add ability to reduce the Rx latency
On 14/09/2021 02:31, Alvin Zhang wrote:
This patch adds a devarg parameter to enable/disable reducing the Rx
latency.
Signed-off-by: Alvin Zhang
---
doc/guides/nics/ice.rst | 8
drivers/net/ice/ice_ethdev.c | 26 +++---
drivers/net/ice/ice_ethdev.h | 1 +
3 files changed, 32 insertions(+), 3 deletions(-)
diff --git a/doc/guides/nics/ice.rst b/doc/guides/nics/ice.rst index
5bc472f..3db0430 100644
--- a/doc/guides/nics/ice.rst
+++ b/doc/guides/nics/ice.rst
@@ -219,6 +219,14 @@ Runtime Config Options
These ICE_DBG_XXX are defined in ``drivers/net/ice/base/ice_type.h``.
+- ``Reduce Rx interrupts and latency`` (default ``0``)
+
+ vRAN workloads require low latency DPDK interface for the front
+ haul interface connection to Radio. Now we can reduce Rx interrupts
+ and latency by specify ``1`` for parameter ``rx-low-latency``::
+
+-a :88:00.0,rx-low-latency=1
+
When would a user select this and when not? What is the trade off?
The text is a bit unclear. It looks below like it reduces the interrupt
latency, but
not the number of interrupts. Maybe I got it wrong.
Yes, it reduces the interrupt latency,
We will refine the doc in next patch.
Thanks, the text in v2 is clearer.
Driver compilation and testing
--
diff --git a/drivers/net/ice/ice_ethdev.c
b/drivers/net/ice/ice_ethdev.c index a4cd39c..85662e4 100644
--- a/drivers/net/ice/ice_ethdev.c
+++ b/drivers/net/ice/ice_ethdev.c
@@ -29,12 +29,14 @@
#define ICE_PIPELINE_MODE_SUPPORT_ARG "pipeline-mode-support"
#define ICE_PROTO_XTR_ARG "proto_xtr"
#define ICE_HW_DEBUG_MASK_ARG "hw_debug_mask"
+#define ICE_RX_LOW_LATENCY"rx-low-latency"
static const char * const ice_valid_args[] = {
ICE_SAFE_MODE_SUPPORT_ARG,
ICE_PIPELINE_MODE_SUPPORT_ARG,
ICE_PROTO_XTR_ARG,
ICE_HW_DEBUG_MASK_ARG,
+ ICE_RX_LOW_LATENCY,
NULL
};
@@ -1827,6 +1829,9 @@ static int ice_parse_devargs(struct rte_eth_dev
*dev)
if (ret)
goto bail;
+ ret = rte_kvargs_process(kvlist, ICE_RX_LOW_LATENCY,
+&parse_bool, &ad->devargs.rx_low_latency);
+
bail:
rte_kvargs_free(kvlist);
return ret;
@@ -3144,8 +3149,9 @@ static int ice_init_rss(struct ice_pf *pf) {
struct ice_hw *hw = ICE_VSI_TO_HW(vsi);
uint32_t val, val_tx;
- int i;
+ int rx_low_latency, i;
+ rx_low_latency = vsi->adapter->devargs.rx_low_latency;
for (i = 0; i < nb_queue; i++) {
/*do actual bind*/
val = (msix_vect & QINT_RQCTL_MSIX_INDX_M) | @@ -3155,8
+3161,21 @@
static int ice_init_rss(struct ice_pf *pf)
PMD_DRV_LOG(INFO, "queue %d is binding to vect %d",
base_queue + i, msix_vect);
+
/* set ITR0 value */
- ICE_WRITE_REG(hw, GLINT_ITR(0, msix_vect), 0x2);
+ if (rx_low_latency) {
+ /**
+* Empirical configuration for optimal real time
+* latency reduced interrupt throttling to 2us
+*/
+ ICE_WRITE_REG(hw, GLINT_ITR(0, msix_vect), 0x1);
Why not set this to 0? "Setting the INTERVAL to zero enables immediate
interrupt."
Didn't see a reply to this comment?
I'm not requesting a change, just asking if there is a reason you didn't
choose the lowest latency setting, and if you should?
+ ICE_WRITE_REG(hw, QRX_ITR(base_queue + i),
+ QRX_ITR_NO_EXPR_M);
+ } else {
+ ICE_WRITE_REG(hw, GLINT_ITR(0, msix_vect), 0x2);
+ ICE_WRITE_REG(hw, QRX_ITR(base_queue + i), 0);
+ }
+
ICE_WRITE_REG(hw, QINT_RQCTL(base_queue + i), val);
ICE_WRITE_REG(hw, QINT_TQCTL(base_queue + i), val_tx);
}
@@ -5314,7 +5333,8 @@ static int ice_xstats_get_names(__rte_unused
struct rte_eth_dev *dev,
ICE_HW_DEBUG_MASK_ARG "=0xXXX"
ICE_PROTO_XTR_ARG
"=[queue:]"
ICE_SAFE_MODE_SUPPORT_ARG "=<0|1>"
- ICE_PIPELINE_MODE_SUPPORT_ARG "=<0|1>");
+ ICE_PIPELINE_MODE_SUPPORT_ARG "=<0|1>"
+ ICE_RX_LOW_LATENCY "=<0|1>");
RTE_LOG_REGISTER_SUFFIX(ice_logtype_init, init, NOTICE);
RTE_LOG_REGISTER_SUFFIX(ice_logtype_driver, driver, NOTICE); diff
--git a/drivers/net/ice/ice_ethdev.h b/drivers/net/ice/ice_ethdev.h
index b4bf651..c61cc1f 100644
--- a/dr
Re: [dpdk-dev] [PATCH v2 0/4] delete HW rings when releasing queues for some drivers
> -Original Message-
> From: David Marchand
> Sent: Monday, September 20, 2021 15:25
> To: Yunjian Wang ; Wang, Haiyue
> ; Yigit, Ferruh
>
> Cc: dev ; Xing, Beilei ; Yang, Qiming
> ;
> Zhang, Qi Z ; dingxiaoxi...@huawei.com; Renata Saiakhova
>
> Subject: Re: [dpdk-dev] [PATCH v2 0/4] delete HW rings when releasing queues
> for some drivers
>
> On Sat, Sep 18, 2021 at 10:34 AM Yunjian Wang wrote:
> >
> > This series for deleting HW rings when releasing queues for
> > igb, ixgbe, i40e, ice & em drivers.
> >
> > ---
> > v2:
> >* Update commit log
> >
> > Yunjian Wang (4):
> > net/e1000: delete HW rings when releasing queues
> > net/ice: delete HW rings when releasing queues
> > net/i40e: delete HW rings when releasing queues
> > net/ixgbe: delete HW rings when releasing queues
> >
> > drivers/net/e1000/em_rxtx.c| 8 ++--
> > drivers/net/e1000/igb_rxtx.c | 9 +++--
> > drivers/net/i40e/i40e_fdir.c | 3 ---
> > drivers/net/i40e/i40e_rxtx.c | 8 ++--
> > drivers/net/i40e/i40e_rxtx.h | 2 ++
> > drivers/net/ice/ice_rxtx.c | 6 --
> > drivers/net/ice/ice_rxtx.h | 2 ++
> > drivers/net/ixgbe/ixgbe_rxtx.c | 6 --
> > drivers/net/ixgbe/ixgbe_rxtx.h | 2 ++
> > 9 files changed, 33 insertions(+), 13 deletions(-)
> >
>
> - In net/ice (at least), the fdir rxq/txq memzones can be aligned on
> the same scheme.
> Looking at the remaining drivers (net/cnxk, net/cxgbe and
> net/octeontx2), we could apply the same principle of keeping a
> reference to mz in internal driver structures.
> Afterwards, I see no need to keep rte_eth_dma_zone_free() (it's
> internal, so we can remove and it's easy to re-add if a need arises).
>
> Wdyt?
Yes, this makes dma_zone management clean. And good for hotplug design.
Then the function 'rte_eth_dma_zone_reserve' can also be simplified, no
need to lookup and check, just call 'rte_memzone_reserve_aligned' directly ?
mz = rte_memzone_lookup(z_name);
if (mz) {
if ((socket_id != SOCKET_ID_ANY && socket_id != mz->socket_id)
||
size > mz->len ||
((uintptr_t)mz->addr & (align - 1)) != 0) {
RTE_ETHDEV_LOG(ERR,
"memzone %s does not justify the requested
attributes\n",
mz->name);
return NULL;
}
return mz;
}
>
>
> - Is this worth backporting to stable branches?
>
+1
>
> --
> David Marchand
[dpdk-dev] [PATCH v2 2/5] test/event: add unit test for event buffer size config api
this patch adds unit test for rte_event_eth_rx_adapter_create_with_params
api and validate all possible input combinations.
Signed-off-by: Naga Harish K S V
---
app/test/test_event_eth_rx_adapter.c | 53 +---
1 file changed, 49 insertions(+), 4 deletions(-)
diff --git a/app/test/test_event_eth_rx_adapter.c
b/app/test/test_event_eth_rx_adapter.c
index add4d8a678..3c0f0ad7cc 100644
--- a/app/test/test_event_eth_rx_adapter.c
+++ b/app/test/test_event_eth_rx_adapter.c
@@ -428,6 +428,50 @@ adapter_create_free(void)
return TEST_SUCCESS;
}
+static int
+adapter_create_free_v2(void)
+{
+ int err;
+
+ struct rte_event_port_conf rx_p_conf = {
+ .dequeue_depth = 8,
+ .enqueue_depth = 8,
+ .new_event_threshold = 1200,
+ };
+
+ struct rte_event_eth_rx_adapter_params rxa_params = {
+ .event_buf_size = 1024
+ };
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, NULL, NULL);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EEXIST, "Expected -EEXIST %d got %d", -EEXIST, err);
+
+ rxa_params.event_buf_size = 0;
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ err = rte_event_eth_rx_adapter_free(TEST_INST_ID);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_free(TEST_INST_ID);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL %d got %d", -EINVAL, err);
+
+ err = rte_event_eth_rx_adapter_free(1);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL %d got %d", -EINVAL, err);
+
+ return TEST_SUCCESS;
+}
+
static int
adapter_queue_add_del(void)
{
@@ -435,7 +479,7 @@ adapter_queue_add_del(void)
struct rte_event ev;
uint32_t cap;
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
err = rte_event_eth_rx_adapter_caps_get(TEST_DEV_ID, TEST_ETHDEV_ID,
&cap);
@@ -523,7 +567,7 @@ adapter_multi_eth_add_del(void)
uint16_t port_index, port_index_base, drv_id = 0;
char driver_name[50];
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
ev.queue_id = 0;
ev.sched_type = RTE_SCHED_TYPE_ATOMIC;
@@ -594,7 +638,7 @@ adapter_intr_queue_add_del(void)
struct rte_event ev;
uint32_t cap;
uint16_t eth_port;
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
if (!default_params.rx_intr_port_inited)
return 0;
@@ -687,7 +731,7 @@ adapter_start_stop(void)
ev.sched_type = RTE_SCHED_TYPE_ATOMIC;
ev.priority = 0;
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
queue_config.rx_queue_flags = 0;
if (default_params.caps &
@@ -753,6 +797,7 @@ static struct unit_test_suite event_eth_rx_tests = {
.teardown = testsuite_teardown,
.unit_test_cases = {
TEST_CASE_ST(NULL, NULL, adapter_create_free),
+ TEST_CASE_ST(NULL, NULL, adapter_create_free_v2),
TEST_CASE_ST(adapter_create, adapter_free,
adapter_queue_add_del),
TEST_CASE_ST(adapter_create, adapter_free,
--
2.25.1
[dpdk-dev] [PATCH v2 3/5] eventdev/rx_adapter:add per queue event buffer configure support
To configure per queue event buffer size, applications sets
``rte_event_eth_rx_adapter_params::use_queue_event_buf`` flag
as true and is passed to ``rte_event_eth_rx_adapter_create_with_params``
api.
The per queue event buffer size is populated in
``rte_event_eth_rx_adapter_queue_conf::event_buf_size`` and passed
to ``rte_event_eth_rx_adapter_queue_add`` api.
Signed-off-by: Naga Harish K S V
---
.../prog_guide/event_ethernet_rx_adapter.rst | 19 ---
lib/eventdev/rte_event_eth_rx_adapter.h | 4
2 files changed, 16 insertions(+), 7 deletions(-)
diff --git a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
index dd753613bd..333e6f8192 100644
--- a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
+++ b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
@@ -62,12 +62,14 @@ service function and needs to create an event port for it.
The callback is
expected to fill the ``struct rte_event_eth_rx_adapter_conf structure``
passed to it.
-If the application desires to control the event buffer size, it can use the
-``rte_event_eth_rx_adapter_create_with_params()`` api. The event buffer size is
-specified using ``struct rte_event_eth_rx_adapter_params::event_buf_size``.
-The function is passed the event device to be associated with the adapter
-and port configuration for the adapter to setup an event port if the
-adapter needs to use a service function.
+If the application desires to control the event buffer size at adapter level,
+it can use the ``rte_event_eth_rx_adapter_create_with_params()`` api. The event
+buffer size is specified using ``struct rte_event_eth_rx_adapter_params::
+event_buf_size``. To configure the event buffer size at queue level, the
boolean
+flag ``struct rte_event_eth_rx_adapter_params::use_queue_event_buf`` need to be
+set to true. The function is passed the event device to be associated with
+the adapter and port configuration for the adapter to setup an event port
+if the adapter needs to use a service function.
Adding Rx Queues to the Adapter Instance
@@ -79,7 +81,9 @@ parameter. Event information for packets from this Rx queue
is encoded in the
``ev`` field of ``struct rte_event_eth_rx_adapter_queue_conf``. The
servicing_weight member of the struct rte_event_eth_rx_adapter_queue_conf
is the relative polling frequency of the Rx queue and is applicable when the
-adapter uses a service core function.
+adapter uses a service core function. The applications can configure queue
+event buffer size in ``struct
rte_event_eth_rx_adapter_queue_conf::event_buf_size``
+parameter.
.. code-block:: c
@@ -90,6 +94,7 @@ adapter uses a service core function.
queue_config.rx_queue_flags = 0;
queue_config.ev = ev;
queue_config.servicing_weight = 1;
+ queue_config.event_buf_size = 1024;
err = rte_event_eth_rx_adapter_queue_add(id,
eth_dev_id,
diff --git a/lib/eventdev/rte_event_eth_rx_adapter.h
b/lib/eventdev/rte_event_eth_rx_adapter.h
index a7881097b4..b9f0563244 100644
--- a/lib/eventdev/rte_event_eth_rx_adapter.h
+++ b/lib/eventdev/rte_event_eth_rx_adapter.h
@@ -199,6 +199,8 @@ struct rte_event_eth_rx_adapter_queue_conf {
* Valid when RTE_EVENT_ETH_RX_ADAPTER_QUEUE_EVENT_VECTOR flag is set in
* @see rte_event_eth_rx_adapter_queue_conf::rx_queue_flags.
*/
+ uint16_t event_buf_size;
+ /**< event buffer size for this queue */
};
/**
@@ -265,6 +267,8 @@ struct rte_event_eth_rx_adapter_params {
/**< size of event buffer for the adapter.
* the size is aligned to BATCH_SIZE and added (2 * BATCH_SIZE)
*/
+ bool use_queue_event_buf;
+ /**< flag to indicate that event buffer is separate for each queue */
};
/**
--
2.25.1
[dpdk-dev] [PATCH v2 1/5] eventdev/rx_adapter: add support to configure event buffer size
Currently Rx event buffer is static array with a default size
of 192(6*BATCH_SIZE).
``rte_event_eth_rx_adapter_create_with_params`` api is added which takes
``struct rte_event_eth_rx_adapter_params`` to configure event
buffer size in addition other params . The event buffer is
allocated from heap after aligning the size to BATCH_SIZE and
adding 2*BATCH_SIZE. In case of NULL params argument, default event
buffer size is used.
Signed-off-by: Naga Harish K S V
Signed-off-by: Ganapati Kundapura
---
.../prog_guide/event_ethernet_rx_adapter.rst | 7 ++
lib/eventdev/rte_event_eth_rx_adapter.c | 94 +--
lib/eventdev/rte_event_eth_rx_adapter.h | 40 +++-
lib/eventdev/version.map | 2 +
4 files changed, 135 insertions(+), 8 deletions(-)
diff --git a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
index 0780b6f711..dd753613bd 100644
--- a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
+++ b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
@@ -62,6 +62,13 @@ service function and needs to create an event port for it.
The callback is
expected to fill the ``struct rte_event_eth_rx_adapter_conf structure``
passed to it.
+If the application desires to control the event buffer size, it can use the
+``rte_event_eth_rx_adapter_create_with_params()`` api. The event buffer size is
+specified using ``struct rte_event_eth_rx_adapter_params::event_buf_size``.
+The function is passed the event device to be associated with the adapter
+and port configuration for the adapter to setup an event port if the
+adapter needs to use a service function.
+
Adding Rx Queues to the Adapter Instance
diff --git a/lib/eventdev/rte_event_eth_rx_adapter.c
b/lib/eventdev/rte_event_eth_rx_adapter.c
index f2dc69503d..df1653b497 100644
--- a/lib/eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/eventdev/rte_event_eth_rx_adapter.c
@@ -82,7 +82,9 @@ struct rte_eth_event_enqueue_buffer {
/* Count of events in this buffer */
uint16_t count;
/* Array of events in this buffer */
- struct rte_event events[ETH_EVENT_BUFFER_SIZE];
+ struct rte_event *events;
+ /* size of event buffer */
+ uint16_t events_size;
/* Event enqueue happens from head */
uint16_t head;
/* New packets from rte_eth_rx_burst is enqued from tail */
@@ -919,7 +921,7 @@ rxa_buffer_mbufs(struct rte_event_eth_rx_adapter
*rx_adapter,
dropped = 0;
nb_cb = dev_info->cb_fn(eth_dev_id, rx_queue_id,
buf->last |
- (RTE_DIM(buf->events) & ~buf->last_mask),
+ (buf->events_size & ~buf->last_mask),
buf->count >= BATCH_SIZE ?
buf->count - BATCH_SIZE : 0,
&buf->events[buf->tail],
@@ -945,7 +947,7 @@ rxa_pkt_buf_available(struct rte_eth_event_enqueue_buffer
*buf)
uint32_t nb_req = buf->tail + BATCH_SIZE;
if (!buf->last) {
- if (nb_req <= RTE_DIM(buf->events))
+ if (nb_req <= buf->events_size)
return true;
if (buf->head >= BATCH_SIZE) {
@@ -2164,12 +2166,15 @@ rxa_ctrl(uint8_t id, int start)
return 0;
}
-int
-rte_event_eth_rx_adapter_create_ext(uint8_t id, uint8_t dev_id,
- rte_event_eth_rx_adapter_conf_cb conf_cb,
- void *conf_arg)
+static int
+rxa_create(uint8_t id, uint8_t dev_id,
+ struct rte_event_eth_rx_adapter_params *rxa_params,
+ rte_event_eth_rx_adapter_conf_cb conf_cb,
+ void *conf_arg)
{
struct rte_event_eth_rx_adapter *rx_adapter;
+ struct rte_eth_event_enqueue_buffer *buf;
+ struct rte_event *events;
int ret;
int socket_id;
uint16_t i;
@@ -2184,6 +2189,7 @@ rte_event_eth_rx_adapter_create_ext(uint8_t id, uint8_t
dev_id,
RTE_EVENT_ETH_RX_ADAPTER_ID_VALID_OR_ERR_RET(id, -EINVAL);
RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
+
if (conf_cb == NULL)
return -EINVAL;
@@ -2231,11 +2237,30 @@ rte_event_eth_rx_adapter_create_ext(uint8_t id, uint8_t
dev_id,
rte_free(rx_adapter);
return -ENOMEM;
}
+
rte_spinlock_init(&rx_adapter->rx_lock);
+
for (i = 0; i < RTE_MAX_ETHPORTS; i++)
rx_adapter->eth_devices[i].dev = &rte_eth_devices[i];
+ /* Rx adapter event buffer allocation */
+ buf = &rx_adapter->event_enqueue_buffer;
+ buf->events_size = RTE_ALIGN(rxa_params->event_buf_size, BATCH_SIZE);
+
+ events = rte_zmalloc_socket(rx_adapter->mem_name,
+ buf->events_size * sizeof(*events),
+
[dpdk-dev] [PATCH v2 4/5] eventdev/rx_adapter: implement per queue event buffer
this patch implement the per queue event buffer after
required validations.
Signed-off-by: Naga Harish K S V
---
lib/eventdev/rte_event_eth_rx_adapter.c | 188 ++--
1 file changed, 139 insertions(+), 49 deletions(-)
diff --git a/lib/eventdev/rte_event_eth_rx_adapter.c
b/lib/eventdev/rte_event_eth_rx_adapter.c
index df1653b497..20ea440275 100644
--- a/lib/eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/eventdev/rte_event_eth_rx_adapter.c
@@ -99,10 +99,12 @@ struct rte_event_eth_rx_adapter {
uint8_t rss_key_be[RSS_KEY_SIZE];
/* Event device identifier */
uint8_t eventdev_id;
- /* Per ethernet device structure */
- struct eth_device_info *eth_devices;
/* Event port identifier */
uint8_t event_port_id;
+ /* Flag indicating per rxq event buffer */
+ bool use_queue_event_buf;
+ /* Per ethernet device structure */
+ struct eth_device_info *eth_devices;
/* Lock to serialize config updates with service function */
rte_spinlock_t rx_lock;
/* Max mbufs processed in any service function invocation */
@@ -238,6 +240,7 @@ struct eth_rx_queue_info {
uint32_t flow_id_mask; /* Set to ~0 if app provides flow id else 0 */
uint64_t event;
struct eth_rx_vector_data vector_data;
+ struct rte_eth_event_enqueue_buffer *event_buf;
};
static struct rte_event_eth_rx_adapter **event_eth_rx_adapter;
@@ -753,10 +756,9 @@ rxa_enq_block_end_ts(struct rte_event_eth_rx_adapter
*rx_adapter,
/* Enqueue buffered events to event device */
static inline uint16_t
-rxa_flush_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter)
+rxa_flush_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter,
+ struct rte_eth_event_enqueue_buffer *buf)
{
- struct rte_eth_event_enqueue_buffer *buf =
- &rx_adapter->event_enqueue_buffer;
struct rte_event_eth_rx_adapter_stats *stats = &rx_adapter->stats;
uint16_t count = buf->last ? buf->last - buf->head : buf->count;
@@ -874,15 +876,14 @@ rxa_buffer_mbufs(struct rte_event_eth_rx_adapter
*rx_adapter,
uint16_t eth_dev_id,
uint16_t rx_queue_id,
struct rte_mbuf **mbufs,
- uint16_t num)
+ uint16_t num,
+ struct rte_eth_event_enqueue_buffer *buf)
{
uint32_t i;
struct eth_device_info *dev_info =
&rx_adapter->eth_devices[eth_dev_id];
struct eth_rx_queue_info *eth_rx_queue_info =
&dev_info->rx_queue[rx_queue_id];
- struct rte_eth_event_enqueue_buffer *buf =
- &rx_adapter->event_enqueue_buffer;
uint16_t new_tail = buf->tail;
uint64_t event = eth_rx_queue_info->event;
uint32_t flow_id_mask = eth_rx_queue_info->flow_id_mask;
@@ -968,11 +969,10 @@ rxa_eth_rx(struct rte_event_eth_rx_adapter *rx_adapter,
uint16_t queue_id,
uint32_t rx_count,
uint32_t max_rx,
- int *rxq_empty)
+ int *rxq_empty,
+ struct rte_eth_event_enqueue_buffer *buf)
{
struct rte_mbuf *mbufs[BATCH_SIZE];
- struct rte_eth_event_enqueue_buffer *buf =
- &rx_adapter->event_enqueue_buffer;
struct rte_event_eth_rx_adapter_stats *stats =
&rx_adapter->stats;
uint16_t n;
@@ -985,7 +985,7 @@ rxa_eth_rx(struct rte_event_eth_rx_adapter *rx_adapter,
*/
while (rxa_pkt_buf_available(buf)) {
if (buf->count >= BATCH_SIZE)
- rxa_flush_event_buffer(rx_adapter);
+ rxa_flush_event_buffer(rx_adapter, buf);
stats->rx_poll_count++;
n = rte_eth_rx_burst(port_id, queue_id, mbufs, BATCH_SIZE);
@@ -994,14 +994,14 @@ rxa_eth_rx(struct rte_event_eth_rx_adapter *rx_adapter,
*rxq_empty = 1;
break;
}
- rxa_buffer_mbufs(rx_adapter, port_id, queue_id, mbufs, n);
+ rxa_buffer_mbufs(rx_adapter, port_id, queue_id, mbufs, n, buf);
nb_rx += n;
if (rx_count + nb_rx > max_rx)
break;
}
if (buf->count > 0)
- rxa_flush_event_buffer(rx_adapter);
+ rxa_flush_event_buffer(rx_adapter, buf);
return nb_rx;
}
@@ -1142,7 +1142,7 @@ rxa_intr_ring_dequeue(struct rte_event_eth_rx_adapter
*rx_adapter)
ring_lock = &rx_adapter->intr_ring_lock;
if (buf->count >= BATCH_SIZE)
- rxa_flush_event_buffer(rx_adapter);
+ rxa_flush_event_buffer(rx_adapter, buf);
while (rxa_pkt_buf_available(buf)) {
struct eth_device_info *dev_info;
@@ -1194,7 +1194,7 @@ rxa_intr_ring_dequeue(struct rte_event_eth_rx_adapter
*rx
[dpdk-dev] [PATCH v2 5/5] test/eventdev: add per rx queue event buffer unit
this patch adds unit tests for per rx queue event buffer
Signed-off-by: Naga Harish K S V
---
app/test/test_event_eth_rx_adapter.c | 86
1 file changed, 86 insertions(+)
diff --git a/app/test/test_event_eth_rx_adapter.c
b/app/test/test_event_eth_rx_adapter.c
index 3c0f0ad7cc..0564d0 100644
--- a/app/test/test_event_eth_rx_adapter.c
+++ b/app/test/test_event_eth_rx_adapter.c
@@ -387,6 +387,90 @@ adapter_create(void)
return err;
}
+static int
+adapter_create_with_params(void)
+{
+ int err;
+ struct rte_event_dev_info dev_info;
+ struct rte_event_port_conf rx_p_conf;
+ struct rte_event_eth_rx_adapter_params rxa_params;
+
+ memset(&rx_p_conf, 0, sizeof(rx_p_conf));
+
+ err = rte_event_dev_info_get(TEST_DEV_ID, &dev_info);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ rx_p_conf.new_event_threshold = dev_info.max_num_events;
+ rx_p_conf.dequeue_depth = dev_info.max_event_port_dequeue_depth;
+ rx_p_conf.enqueue_depth = dev_info.max_event_port_enqueue_depth;
+
+ rxa_params.use_queue_event_buf = false;
+ rxa_params.event_buf_size = 0;
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ rxa_params.use_queue_event_buf = true;
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EEXIST, "Expected -EEXIST got %d", err);
+
+ return TEST_SUCCESS;
+}
+
+static int
+adapter_queue_event_buf_test(void)
+{
+ int err;
+ struct rte_event ev;
+ uint32_t cap;
+
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
+
+ err = rte_event_eth_rx_adapter_caps_get(TEST_DEV_ID, TEST_ETHDEV_ID,
+&cap);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ ev.queue_id = 0;
+ ev.sched_type = RTE_SCHED_TYPE_ATOMIC;
+ ev.priority = 0;
+
+ queue_config.rx_queue_flags = 0;
+ if (cap & RTE_EVENT_ETH_RX_ADAPTER_CAP_OVERRIDE_FLOW_ID) {
+ ev.flow_id = 1;
+ queue_config.rx_queue_flags =
+ RTE_EVENT_ETH_RX_ADAPTER_QUEUE_FLOW_ID_VALID;
+ }
+ queue_config.ev = ev;
+ queue_config.servicing_weight = 1;
+ queue_config.event_buf_size = 0;
+
+ err = rte_event_eth_rx_adapter_queue_add(TEST_INST_ID,
+ TEST_ETHDEV_ID, 0,
+ &queue_config);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ queue_config.event_buf_size = 1024;
+
+ err = rte_event_eth_rx_adapter_queue_add(TEST_INST_ID,
+ TEST_ETHDEV_ID, 0,
+ &queue_config);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_queue_del(TEST_INST_ID,
+ TEST_ETHDEV_ID,
+ 0);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ return TEST_SUCCESS;
+}
+
static void
adapter_free(void)
{
@@ -804,6 +888,8 @@ static struct unit_test_suite event_eth_rx_tests = {
adapter_multi_eth_add_del),
TEST_CASE_ST(adapter_create, adapter_free, adapter_start_stop),
TEST_CASE_ST(adapter_create, adapter_free, adapter_stats),
+ TEST_CASE_ST(adapter_create_with_params, adapter_free,
+adapter_queue_event_buf_test),
TEST_CASES_END() /**< NULL terminate unit test array */
}
};
--
2.25.1
Re: [dpdk-dev] [dpdk-stable] [PATCH] net: fix checksum API documentation
On 9/17/2021 11:57 AM, Morten Brørup wrote:
>> From: Lance Richardson [mailto:lance.richard...@broadcom.com]
>> Sent: Thursday, 16 September 2021 18.11
>>
>> Minor corrections and improvements to documentation
>> for checksum APIs.
>>
>> Fixes: 6006818cfb26 ("net: new checksum functions")
>> Fixes: 45a08ef55e44 ("net: introduce functions to verify L4 checksums")
>> Cc: sta...@dpdk.org
>> Signed-off-by: Lance Richardson
>> ---
>> lib/net/rte_ip.h | 8
>> 1 file changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/lib/net/rte_ip.h b/lib/net/rte_ip.h
>> index 05948b69b7..fd08ea31b2 100644
>> --- a/lib/net/rte_ip.h
>> +++ b/lib/net/rte_ip.h
>> @@ -488,7 +488,7 @@ rte_ipv6_phdr_cksum(const struct rte_ipv6_hdr
>> *ipv6_hdr, uint64_t ol_flags)
>> }
>>
>> /**
>> - * @internal Calculate the non-complemented IPv4 L4 checksum
>> + * @internal Calculate the non-complemented IPv6 L4 checksum
>> */
>> static inline uint16_t
>> __rte_ipv6_udptcp_cksum(const struct rte_ipv6_hdr *ipv6_hdr, const
>> void *l4_hdr)
>> @@ -509,15 +509,15 @@ __rte_ipv6_udptcp_cksum(const struct rte_ipv6_hdr
>> *ipv6_hdr, const void *l4_hdr)
>> /**
>> * Process the IPv6 UDP or TCP checksum.
>> *
>> - * The IPv4 header should not contains options. The layer 4 checksum
>> - * must be set to 0 in the packet by the caller.
>> + * The IPv6 header must not be followed by extension headers. The
>> layer 4
>> + * checksum must be set to 0 in the L4 header by the caller.
>> *
>> * @param ipv6_hdr
>> * The pointer to the contiguous IPv6 header.
>> * @param l4_hdr
>> * The pointer to the beginning of the L4 header.
>> * @return
>> - * The complemented checksum to set in the IP packet.
>> + * The complemented checksum to set in the L4 header.
Isn't this wrong for 'rte_ipv4_udptcp_cksum()' too? Since you are touching this,
can you fix that one too?
>> */
>> static inline uint16_t
>> rte_ipv6_udptcp_cksum(const struct rte_ipv6_hdr *ipv6_hdr, const void
>> *l4_hdr)
>> --
>> 2.25.1
>
> Reviewed-by: Morten Brørup
>
[dpdk-dev] [PATCH v2 2/5] test/event: add unit test for event buffer size config api
this patch adds unit test for rte_event_eth_rx_adapter_create_with_params
api and validate all possible input combinations.
Signed-off-by: Naga Harish K S V
---
app/test/test_event_eth_rx_adapter.c | 53 +---
1 file changed, 49 insertions(+), 4 deletions(-)
diff --git a/app/test/test_event_eth_rx_adapter.c
b/app/test/test_event_eth_rx_adapter.c
index add4d8a678..3c0f0ad7cc 100644
--- a/app/test/test_event_eth_rx_adapter.c
+++ b/app/test/test_event_eth_rx_adapter.c
@@ -428,6 +428,50 @@ adapter_create_free(void)
return TEST_SUCCESS;
}
+static int
+adapter_create_free_v2(void)
+{
+ int err;
+
+ struct rte_event_port_conf rx_p_conf = {
+ .dequeue_depth = 8,
+ .enqueue_depth = 8,
+ .new_event_threshold = 1200,
+ };
+
+ struct rte_event_eth_rx_adapter_params rxa_params = {
+ .event_buf_size = 1024
+ };
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, NULL, NULL);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EEXIST, "Expected -EEXIST %d got %d", -EEXIST, err);
+
+ rxa_params.event_buf_size = 0;
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ err = rte_event_eth_rx_adapter_free(TEST_INST_ID);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_free(TEST_INST_ID);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL %d got %d", -EINVAL, err);
+
+ err = rte_event_eth_rx_adapter_free(1);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL %d got %d", -EINVAL, err);
+
+ return TEST_SUCCESS;
+}
+
static int
adapter_queue_add_del(void)
{
@@ -435,7 +479,7 @@ adapter_queue_add_del(void)
struct rte_event ev;
uint32_t cap;
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
err = rte_event_eth_rx_adapter_caps_get(TEST_DEV_ID, TEST_ETHDEV_ID,
&cap);
@@ -523,7 +567,7 @@ adapter_multi_eth_add_del(void)
uint16_t port_index, port_index_base, drv_id = 0;
char driver_name[50];
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
ev.queue_id = 0;
ev.sched_type = RTE_SCHED_TYPE_ATOMIC;
@@ -594,7 +638,7 @@ adapter_intr_queue_add_del(void)
struct rte_event ev;
uint32_t cap;
uint16_t eth_port;
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
if (!default_params.rx_intr_port_inited)
return 0;
@@ -687,7 +731,7 @@ adapter_start_stop(void)
ev.sched_type = RTE_SCHED_TYPE_ATOMIC;
ev.priority = 0;
- struct rte_event_eth_rx_adapter_queue_conf queue_config;
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
queue_config.rx_queue_flags = 0;
if (default_params.caps &
@@ -753,6 +797,7 @@ static struct unit_test_suite event_eth_rx_tests = {
.teardown = testsuite_teardown,
.unit_test_cases = {
TEST_CASE_ST(NULL, NULL, adapter_create_free),
+ TEST_CASE_ST(NULL, NULL, adapter_create_free_v2),
TEST_CASE_ST(adapter_create, adapter_free,
adapter_queue_add_del),
TEST_CASE_ST(adapter_create, adapter_free,
--
2.25.1
[dpdk-dev] [PATCH v2 3/5] eventdev/rx_adapter:add per queue event buffer configure support
To configure per queue event buffer size, applications sets
``rte_event_eth_rx_adapter_params::use_queue_event_buf`` flag
as true and is passed to ``rte_event_eth_rx_adapter_create_with_params``
api.
The per queue event buffer size is populated in
``rte_event_eth_rx_adapter_queue_conf::event_buf_size`` and passed
to ``rte_event_eth_rx_adapter_queue_add`` api.
Signed-off-by: Naga Harish K S V
---
.../prog_guide/event_ethernet_rx_adapter.rst | 19 ---
lib/eventdev/rte_event_eth_rx_adapter.h | 4
2 files changed, 16 insertions(+), 7 deletions(-)
diff --git a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
index dd753613bd..333e6f8192 100644
--- a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
+++ b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
@@ -62,12 +62,14 @@ service function and needs to create an event port for it.
The callback is
expected to fill the ``struct rte_event_eth_rx_adapter_conf structure``
passed to it.
-If the application desires to control the event buffer size, it can use the
-``rte_event_eth_rx_adapter_create_with_params()`` api. The event buffer size is
-specified using ``struct rte_event_eth_rx_adapter_params::event_buf_size``.
-The function is passed the event device to be associated with the adapter
-and port configuration for the adapter to setup an event port if the
-adapter needs to use a service function.
+If the application desires to control the event buffer size at adapter level,
+it can use the ``rte_event_eth_rx_adapter_create_with_params()`` api. The event
+buffer size is specified using ``struct rte_event_eth_rx_adapter_params::
+event_buf_size``. To configure the event buffer size at queue level, the
boolean
+flag ``struct rte_event_eth_rx_adapter_params::use_queue_event_buf`` need to be
+set to true. The function is passed the event device to be associated with
+the adapter and port configuration for the adapter to setup an event port
+if the adapter needs to use a service function.
Adding Rx Queues to the Adapter Instance
@@ -79,7 +81,9 @@ parameter. Event information for packets from this Rx queue
is encoded in the
``ev`` field of ``struct rte_event_eth_rx_adapter_queue_conf``. The
servicing_weight member of the struct rte_event_eth_rx_adapter_queue_conf
is the relative polling frequency of the Rx queue and is applicable when the
-adapter uses a service core function.
+adapter uses a service core function. The applications can configure queue
+event buffer size in ``struct
rte_event_eth_rx_adapter_queue_conf::event_buf_size``
+parameter.
.. code-block:: c
@@ -90,6 +94,7 @@ adapter uses a service core function.
queue_config.rx_queue_flags = 0;
queue_config.ev = ev;
queue_config.servicing_weight = 1;
+ queue_config.event_buf_size = 1024;
err = rte_event_eth_rx_adapter_queue_add(id,
eth_dev_id,
diff --git a/lib/eventdev/rte_event_eth_rx_adapter.h
b/lib/eventdev/rte_event_eth_rx_adapter.h
index a7881097b4..b9f0563244 100644
--- a/lib/eventdev/rte_event_eth_rx_adapter.h
+++ b/lib/eventdev/rte_event_eth_rx_adapter.h
@@ -199,6 +199,8 @@ struct rte_event_eth_rx_adapter_queue_conf {
* Valid when RTE_EVENT_ETH_RX_ADAPTER_QUEUE_EVENT_VECTOR flag is set in
* @see rte_event_eth_rx_adapter_queue_conf::rx_queue_flags.
*/
+ uint16_t event_buf_size;
+ /**< event buffer size for this queue */
};
/**
@@ -265,6 +267,8 @@ struct rte_event_eth_rx_adapter_params {
/**< size of event buffer for the adapter.
* the size is aligned to BATCH_SIZE and added (2 * BATCH_SIZE)
*/
+ bool use_queue_event_buf;
+ /**< flag to indicate that event buffer is separate for each queue */
};
/**
--
2.25.1
[dpdk-dev] [PATCH v2 1/5] eventdev/rx_adapter: add support to configure event buffer size
Currently Rx event buffer is static array with a default size
of 192(6*BATCH_SIZE).
``rte_event_eth_rx_adapter_create_with_params`` api is added which takes
``struct rte_event_eth_rx_adapter_params`` to configure event
buffer size in addition other params . The event buffer is
allocated from heap after aligning the size to BATCH_SIZE and
adding 2*BATCH_SIZE. In case of NULL params argument, default event
buffer size is used.
Signed-off-by: Naga Harish K S V
Signed-off-by: Ganapati Kundapura
---
v2:
* Updated header file and rx adapter documentation as per review comments.
* new api name is modified as rte_event_eth_rx_adapter_create_with_params
as per review comments.
* rxa_params pointer argument Value NULL is allowed to represent the
default values
v1:
* Initial implementation with documentation and unit tests.
---
.../prog_guide/event_ethernet_rx_adapter.rst | 7 ++
lib/eventdev/rte_event_eth_rx_adapter.c | 94 +--
lib/eventdev/rte_event_eth_rx_adapter.h | 40 +++-
lib/eventdev/version.map | 2 +
4 files changed, 135 insertions(+), 8 deletions(-)
diff --git a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
index 0780b6f711..dd753613bd 100644
--- a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
+++ b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
@@ -62,6 +62,13 @@ service function and needs to create an event port for it.
The callback is
expected to fill the ``struct rte_event_eth_rx_adapter_conf structure``
passed to it.
+If the application desires to control the event buffer size, it can use the
+``rte_event_eth_rx_adapter_create_with_params()`` api. The event buffer size is
+specified using ``struct rte_event_eth_rx_adapter_params::event_buf_size``.
+The function is passed the event device to be associated with the adapter
+and port configuration for the adapter to setup an event port if the
+adapter needs to use a service function.
+
Adding Rx Queues to the Adapter Instance
diff --git a/lib/eventdev/rte_event_eth_rx_adapter.c
b/lib/eventdev/rte_event_eth_rx_adapter.c
index f2dc69503d..df1653b497 100644
--- a/lib/eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/eventdev/rte_event_eth_rx_adapter.c
@@ -82,7 +82,9 @@ struct rte_eth_event_enqueue_buffer {
/* Count of events in this buffer */
uint16_t count;
/* Array of events in this buffer */
- struct rte_event events[ETH_EVENT_BUFFER_SIZE];
+ struct rte_event *events;
+ /* size of event buffer */
+ uint16_t events_size;
/* Event enqueue happens from head */
uint16_t head;
/* New packets from rte_eth_rx_burst is enqued from tail */
@@ -919,7 +921,7 @@ rxa_buffer_mbufs(struct rte_event_eth_rx_adapter
*rx_adapter,
dropped = 0;
nb_cb = dev_info->cb_fn(eth_dev_id, rx_queue_id,
buf->last |
- (RTE_DIM(buf->events) & ~buf->last_mask),
+ (buf->events_size & ~buf->last_mask),
buf->count >= BATCH_SIZE ?
buf->count - BATCH_SIZE : 0,
&buf->events[buf->tail],
@@ -945,7 +947,7 @@ rxa_pkt_buf_available(struct rte_eth_event_enqueue_buffer
*buf)
uint32_t nb_req = buf->tail + BATCH_SIZE;
if (!buf->last) {
- if (nb_req <= RTE_DIM(buf->events))
+ if (nb_req <= buf->events_size)
return true;
if (buf->head >= BATCH_SIZE) {
@@ -2164,12 +2166,15 @@ rxa_ctrl(uint8_t id, int start)
return 0;
}
-int
-rte_event_eth_rx_adapter_create_ext(uint8_t id, uint8_t dev_id,
- rte_event_eth_rx_adapter_conf_cb conf_cb,
- void *conf_arg)
+static int
+rxa_create(uint8_t id, uint8_t dev_id,
+ struct rte_event_eth_rx_adapter_params *rxa_params,
+ rte_event_eth_rx_adapter_conf_cb conf_cb,
+ void *conf_arg)
{
struct rte_event_eth_rx_adapter *rx_adapter;
+ struct rte_eth_event_enqueue_buffer *buf;
+ struct rte_event *events;
int ret;
int socket_id;
uint16_t i;
@@ -2184,6 +2189,7 @@ rte_event_eth_rx_adapter_create_ext(uint8_t id, uint8_t
dev_id,
RTE_EVENT_ETH_RX_ADAPTER_ID_VALID_OR_ERR_RET(id, -EINVAL);
RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
+
if (conf_cb == NULL)
return -EINVAL;
@@ -2231,11 +2237,30 @@ rte_event_eth_rx_adapter_create_ext(uint8_t id, uint8_t
dev_id,
rte_free(rx_adapter);
return -ENOMEM;
}
+
rte_spinlock_init(&rx_adapter->rx_lock);
+
for (i = 0; i < RTE_MAX_ETHPORTS; i++)
rx_adapter->eth_devices[i].dev = &rt
[dpdk-dev] [PATCH v2 4/5] eventdev/rx_adapter: implement per queue event buffer
this patch implement the per queue event buffer after
required validations.
Signed-off-by: Naga Harish K S V
---
lib/eventdev/rte_event_eth_rx_adapter.c | 188 ++--
1 file changed, 139 insertions(+), 49 deletions(-)
diff --git a/lib/eventdev/rte_event_eth_rx_adapter.c
b/lib/eventdev/rte_event_eth_rx_adapter.c
index df1653b497..20ea440275 100644
--- a/lib/eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/eventdev/rte_event_eth_rx_adapter.c
@@ -99,10 +99,12 @@ struct rte_event_eth_rx_adapter {
uint8_t rss_key_be[RSS_KEY_SIZE];
/* Event device identifier */
uint8_t eventdev_id;
- /* Per ethernet device structure */
- struct eth_device_info *eth_devices;
/* Event port identifier */
uint8_t event_port_id;
+ /* Flag indicating per rxq event buffer */
+ bool use_queue_event_buf;
+ /* Per ethernet device structure */
+ struct eth_device_info *eth_devices;
/* Lock to serialize config updates with service function */
rte_spinlock_t rx_lock;
/* Max mbufs processed in any service function invocation */
@@ -238,6 +240,7 @@ struct eth_rx_queue_info {
uint32_t flow_id_mask; /* Set to ~0 if app provides flow id else 0 */
uint64_t event;
struct eth_rx_vector_data vector_data;
+ struct rte_eth_event_enqueue_buffer *event_buf;
};
static struct rte_event_eth_rx_adapter **event_eth_rx_adapter;
@@ -753,10 +756,9 @@ rxa_enq_block_end_ts(struct rte_event_eth_rx_adapter
*rx_adapter,
/* Enqueue buffered events to event device */
static inline uint16_t
-rxa_flush_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter)
+rxa_flush_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter,
+ struct rte_eth_event_enqueue_buffer *buf)
{
- struct rte_eth_event_enqueue_buffer *buf =
- &rx_adapter->event_enqueue_buffer;
struct rte_event_eth_rx_adapter_stats *stats = &rx_adapter->stats;
uint16_t count = buf->last ? buf->last - buf->head : buf->count;
@@ -874,15 +876,14 @@ rxa_buffer_mbufs(struct rte_event_eth_rx_adapter
*rx_adapter,
uint16_t eth_dev_id,
uint16_t rx_queue_id,
struct rte_mbuf **mbufs,
- uint16_t num)
+ uint16_t num,
+ struct rte_eth_event_enqueue_buffer *buf)
{
uint32_t i;
struct eth_device_info *dev_info =
&rx_adapter->eth_devices[eth_dev_id];
struct eth_rx_queue_info *eth_rx_queue_info =
&dev_info->rx_queue[rx_queue_id];
- struct rte_eth_event_enqueue_buffer *buf =
- &rx_adapter->event_enqueue_buffer;
uint16_t new_tail = buf->tail;
uint64_t event = eth_rx_queue_info->event;
uint32_t flow_id_mask = eth_rx_queue_info->flow_id_mask;
@@ -968,11 +969,10 @@ rxa_eth_rx(struct rte_event_eth_rx_adapter *rx_adapter,
uint16_t queue_id,
uint32_t rx_count,
uint32_t max_rx,
- int *rxq_empty)
+ int *rxq_empty,
+ struct rte_eth_event_enqueue_buffer *buf)
{
struct rte_mbuf *mbufs[BATCH_SIZE];
- struct rte_eth_event_enqueue_buffer *buf =
- &rx_adapter->event_enqueue_buffer;
struct rte_event_eth_rx_adapter_stats *stats =
&rx_adapter->stats;
uint16_t n;
@@ -985,7 +985,7 @@ rxa_eth_rx(struct rte_event_eth_rx_adapter *rx_adapter,
*/
while (rxa_pkt_buf_available(buf)) {
if (buf->count >= BATCH_SIZE)
- rxa_flush_event_buffer(rx_adapter);
+ rxa_flush_event_buffer(rx_adapter, buf);
stats->rx_poll_count++;
n = rte_eth_rx_burst(port_id, queue_id, mbufs, BATCH_SIZE);
@@ -994,14 +994,14 @@ rxa_eth_rx(struct rte_event_eth_rx_adapter *rx_adapter,
*rxq_empty = 1;
break;
}
- rxa_buffer_mbufs(rx_adapter, port_id, queue_id, mbufs, n);
+ rxa_buffer_mbufs(rx_adapter, port_id, queue_id, mbufs, n, buf);
nb_rx += n;
if (rx_count + nb_rx > max_rx)
break;
}
if (buf->count > 0)
- rxa_flush_event_buffer(rx_adapter);
+ rxa_flush_event_buffer(rx_adapter, buf);
return nb_rx;
}
@@ -1142,7 +1142,7 @@ rxa_intr_ring_dequeue(struct rte_event_eth_rx_adapter
*rx_adapter)
ring_lock = &rx_adapter->intr_ring_lock;
if (buf->count >= BATCH_SIZE)
- rxa_flush_event_buffer(rx_adapter);
+ rxa_flush_event_buffer(rx_adapter, buf);
while (rxa_pkt_buf_available(buf)) {
struct eth_device_info *dev_info;
@@ -1194,7 +1194,7 @@ rxa_intr_ring_dequeue(struct rte_event_eth_rx_adapter
*rx
[dpdk-dev] [PATCH v2 5/5] test/eventdev: add per rx queue event buffer unit
this patch adds unit tests for per rx queue event buffer
Signed-off-by: Naga Harish K S V
---
app/test/test_event_eth_rx_adapter.c | 86
1 file changed, 86 insertions(+)
diff --git a/app/test/test_event_eth_rx_adapter.c
b/app/test/test_event_eth_rx_adapter.c
index 3c0f0ad7cc..0564d0 100644
--- a/app/test/test_event_eth_rx_adapter.c
+++ b/app/test/test_event_eth_rx_adapter.c
@@ -387,6 +387,90 @@ adapter_create(void)
return err;
}
+static int
+adapter_create_with_params(void)
+{
+ int err;
+ struct rte_event_dev_info dev_info;
+ struct rte_event_port_conf rx_p_conf;
+ struct rte_event_eth_rx_adapter_params rxa_params;
+
+ memset(&rx_p_conf, 0, sizeof(rx_p_conf));
+
+ err = rte_event_dev_info_get(TEST_DEV_ID, &dev_info);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ rx_p_conf.new_event_threshold = dev_info.max_num_events;
+ rx_p_conf.dequeue_depth = dev_info.max_event_port_dequeue_depth;
+ rx_p_conf.enqueue_depth = dev_info.max_event_port_enqueue_depth;
+
+ rxa_params.use_queue_event_buf = false;
+ rxa_params.event_buf_size = 0;
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ rxa_params.use_queue_event_buf = true;
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_create_with_params(TEST_INST_ID,
+ TEST_DEV_ID, &rx_p_conf, &rxa_params);
+ TEST_ASSERT(err == -EEXIST, "Expected -EEXIST got %d", err);
+
+ return TEST_SUCCESS;
+}
+
+static int
+adapter_queue_event_buf_test(void)
+{
+ int err;
+ struct rte_event ev;
+ uint32_t cap;
+
+ struct rte_event_eth_rx_adapter_queue_conf queue_config = {0};
+
+ err = rte_event_eth_rx_adapter_caps_get(TEST_DEV_ID, TEST_ETHDEV_ID,
+&cap);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ ev.queue_id = 0;
+ ev.sched_type = RTE_SCHED_TYPE_ATOMIC;
+ ev.priority = 0;
+
+ queue_config.rx_queue_flags = 0;
+ if (cap & RTE_EVENT_ETH_RX_ADAPTER_CAP_OVERRIDE_FLOW_ID) {
+ ev.flow_id = 1;
+ queue_config.rx_queue_flags =
+ RTE_EVENT_ETH_RX_ADAPTER_QUEUE_FLOW_ID_VALID;
+ }
+ queue_config.ev = ev;
+ queue_config.servicing_weight = 1;
+ queue_config.event_buf_size = 0;
+
+ err = rte_event_eth_rx_adapter_queue_add(TEST_INST_ID,
+ TEST_ETHDEV_ID, 0,
+ &queue_config);
+ TEST_ASSERT(err == -EINVAL, "Expected -EINVAL got %d", err);
+
+ queue_config.event_buf_size = 1024;
+
+ err = rte_event_eth_rx_adapter_queue_add(TEST_INST_ID,
+ TEST_ETHDEV_ID, 0,
+ &queue_config);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ err = rte_event_eth_rx_adapter_queue_del(TEST_INST_ID,
+ TEST_ETHDEV_ID,
+ 0);
+ TEST_ASSERT(err == 0, "Expected 0 got %d", err);
+
+ return TEST_SUCCESS;
+}
+
static void
adapter_free(void)
{
@@ -804,6 +888,8 @@ static struct unit_test_suite event_eth_rx_tests = {
adapter_multi_eth_add_del),
TEST_CASE_ST(adapter_create, adapter_free, adapter_start_stop),
TEST_CASE_ST(adapter_create, adapter_free, adapter_stats),
+ TEST_CASE_ST(adapter_create_with_params, adapter_free,
+adapter_queue_event_buf_test),
TEST_CASES_END() /**< NULL terminate unit test array */
}
};
--
2.25.1
Re: [dpdk-dev] [PATCH v2 1/1] net/e1000: igbvf VLAN offload implementation
> -Original Message-
> From: Renata Saiakhova
> Sent: Wednesday, September 15, 2021 22:52
> To: Wang, Haiyue
> Cc: dev@dpdk.org; Renata Saiakhova
> Subject: [PATCH v2 1/1] net/e1000: igbvf VLAN offload implementation
>
> igbvf_vlan_offload_config and igbvf_vlan_offload_set primal
> implementation, setting vlan filter mask at igbvf_dev_start time.
> Without the above a vlan filter for igbvf is not functional.
>
> Signed-off-by: Renata Saiakhova
> ---
> drivers/net/e1000/igb_ethdev.c | 28
> 1 file changed, 28 insertions(+)
>
> diff --git a/drivers/net/e1000/igb_ethdev.c b/drivers/net/e1000/igb_ethdev.c
> index 10ee0f3341..4c8478427c 100644
> --- a/drivers/net/e1000/igb_ethdev.c
> +++ b/drivers/net/e1000/igb_ethdev.c
> @@ -171,6 +171,8 @@ static int eth_igbvf_xstats_get_names(struct rte_eth_dev
> *dev,
> static int eth_igbvf_stats_reset(struct rte_eth_dev *dev);
> static int igbvf_vlan_filter_set(struct rte_eth_dev *dev,
> uint16_t vlan_id, int on);
> +static int igbvf_vlan_offload_config(struct rte_eth_dev *dev, int mask);
> +static int igbvf_vlan_offload_set(struct rte_eth_dev *dev, int mask);
> static int igbvf_set_vfta(struct e1000_hw *hw, uint16_t vid, bool on);
> static void igbvf_set_vfta_all(struct rte_eth_dev *dev, bool on);
> static int igbvf_default_mac_addr_set(struct rte_eth_dev *dev,
> @@ -410,6 +412,7 @@ static const struct eth_dev_ops igbvf_eth_dev_ops = {
> .xstats_get_names = eth_igbvf_xstats_get_names,
> .stats_reset = eth_igbvf_stats_reset,
> .xstats_reset = eth_igbvf_stats_reset,
> + .vlan_offload_set = igbvf_vlan_offload_set,
> .vlan_filter_set = igbvf_vlan_filter_set,
> .dev_infos_get= eth_igbvf_infos_get,
> .dev_supported_ptypes_get = eth_igb_supported_ptypes_get,
> @@ -3304,6 +3307,8 @@ igbvf_dev_start(struct rte_eth_dev *dev)
> struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
> int ret;
> uint32_t intr_vector = 0;
> + int mask;
> + int err;
>
> PMD_INIT_FUNC_TRACE();
>
> @@ -3313,6 +3318,14 @@ igbvf_dev_start(struct rte_eth_dev *dev)
> /* Set all vfta */
> igbvf_set_vfta_all(dev,1);
>
> + /* Set vlan filter mask */
> + mask = ETH_VLAN_FILTER_MASK;
> + err = igbvf_vlan_offload_config(dev, mask);
> + if (err) {
> + PMD_INIT_LOG(ERR, "Unable to set VLAN offload (%d)", err);
> + return err;
> + }
since igbvf_vlan_offload_config is dummy function, we no need to call
it here.
> +
> eth_igbvf_tx_init(dev);
>
> /* This can fail when allocating mbufs for descriptor rings */
> @@ -3531,6 +3544,21 @@ static void igbvf_set_vfta_all(struct rte_eth_dev
> *dev, bool on)
>
> }
>
> +static int
> +igbvf_vlan_offload_config(__rte_unused struct rte_eth_dev *dev, int mask)
> +{
> + if (mask & ETH_VLAN_STRIP_MASK)
> + return -ENOTSUP;
> + return 0;
> +}
> +
> +static int
> +igbvf_vlan_offload_set(struct rte_eth_dev *dev, int mask)
> +{
> + igbvf_vlan_offload_config(dev, mask);
> + return 0;
> +}
I think we can simplify it to just implement the missed VLAN offload
ops, it will make the VLAN filter by ID to work.
BTW, the API 'rte_eth_dev_set_vlan_offload' will check the mask vs
capabilities, so we can just always "return 0", that means a dummy
function for VF.
static int
eth_igbvf_vlan_offload_set(__rte_unused struct rte_eth_dev *dev,
__rte_unused int mask)
{
return 0;
}
> +
> static int
> igbvf_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on)
> {
> --
> 2.17.2
Re: [dpdk-dev] [PATCH v2 02/15] crypto: add total raw buffer length
Hi Konstantin
On 9/21/2021 12:58 AM, Akhil Goyal wrote:
From: Gagandeep Singh
The current crypto raw data vectors is extended to support
rte_security usecases, where we need total data length to know
how much additional memory space is available in buffer other
than data length so that driver/HW can write expanded size
data after encryption.
Signed-off-by: Gagandeep Singh
Acked-by: Akhil Goyal
---
lib/cryptodev/rte_crypto_sym.h | 6 ++
1 file changed, 6 insertions(+)
diff --git a/lib/cryptodev/rte_crypto_sym.h
b/lib/cryptodev/rte_crypto_sym.h
index dcc0bd5933..e5cef1fb72 100644
--- a/lib/cryptodev/rte_crypto_sym.h
+++ b/lib/cryptodev/rte_crypto_sym.h
@@ -37,6 +37,8 @@ struct rte_crypto_vec {
rte_iova_t iova;
/** length of the data buffer */
uint32_t len;
+ /** total buffer length*/
+ uint32_t tot_len;
};
/**
@@ -980,12 +982,14 @@ rte_crypto_mbuf_to_vec(const struct rte_mbuf
*mb, uint32_t ofs, uint32_t len,
seglen = mb->data_len - ofs;
if (len <= seglen) {
vec[0].len = len;
+ vec[0].tot_len = mb->buf_len;
That doesn't look right.
We should take into a count mbuf headroom and input offset.
Something like:
vec[0].tot_len = mb->buf_len - rte_pktmbuf_headroom(m) - ofs;
Same in other places below.
I believe the packet can expand into headroom based on the protocol support.
Yes, total length is representing the total buffer length available. The
security protocol shall take care of the headroom and offsets.
Re: [dpdk-dev] [PATCH v2 02/15] crypto: add total raw buffer length
> Hi Konstantin
>
> On 9/21/2021 12:58 AM, Akhil Goyal wrote:
> >>> From: Gagandeep Singh
> >>>
> >>> The current crypto raw data vectors is extended to support
> >>> rte_security usecases, where we need total data length to know
> >>> how much additional memory space is available in buffer other
> >>> than data length so that driver/HW can write expanded size
> >>> data after encryption.
> >>>
> >>> Signed-off-by: Gagandeep Singh
> >>> Acked-by: Akhil Goyal
> >>> ---
> >>> lib/cryptodev/rte_crypto_sym.h | 6 ++
> >>> 1 file changed, 6 insertions(+)
> >>>
> >>> diff --git a/lib/cryptodev/rte_crypto_sym.h
> >> b/lib/cryptodev/rte_crypto_sym.h
> >>> index dcc0bd5933..e5cef1fb72 100644
> >>> --- a/lib/cryptodev/rte_crypto_sym.h
> >>> +++ b/lib/cryptodev/rte_crypto_sym.h
> >>> @@ -37,6 +37,8 @@ struct rte_crypto_vec {
> >>> rte_iova_t iova;
> >>> /** length of the data buffer */
> >>> uint32_t len;
> >>> + /** total buffer length*/
> >>> + uint32_t tot_len;
> >>> };
> >>>
> >>> /**
> >>> @@ -980,12 +982,14 @@ rte_crypto_mbuf_to_vec(const struct rte_mbuf
> >> *mb, uint32_t ofs, uint32_t len,
> >>> seglen = mb->data_len - ofs;
> >>> if (len <= seglen) {
> >>> vec[0].len = len;
> >>> + vec[0].tot_len = mb->buf_len;
> >> That doesn't look right.
> >> We should take into a count mbuf headroom and input offset.
> >> Something like:
> >> vec[0].tot_len = mb->buf_len - rte_pktmbuf_headroom(m) - ofs;
> >> Same in other places below.
> >>
> > I believe the packet can expand into headroom based on the protocol support.
> Yes, total length is representing the total buffer length available. The
> security protocol shall take care of the headroom and offsets.
Hmm, and how it will now how many bytes are in head-room, and how many are in
tail-room?
We either need to provide values for both, or assume that only tail-room is
available for the driver.
Re: [dpdk-dev] [dpdk-stable] [PATCH] net: fix checksum API documentation
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Ferruh Yigit
> Sent: Tuesday, 21 September 2021 11.28
>
> On 9/17/2021 11:57 AM, Morten Brørup wrote:
> >> From: Lance Richardson [mailto:lance.richard...@broadcom.com]
> >> Sent: Thursday, 16 September 2021 18.11
> >>
> >> Minor corrections and improvements to documentation
> >> for checksum APIs.
> >>
> >> Fixes: 6006818cfb26 ("net: new checksum functions")
> >> Fixes: 45a08ef55e44 ("net: introduce functions to verify L4
> checksums")
> >> Cc: sta...@dpdk.org
> >> Signed-off-by: Lance Richardson
> >> ---
> >> lib/net/rte_ip.h | 8
> >> 1 file changed, 4 insertions(+), 4 deletions(-)
> >>
> >> diff --git a/lib/net/rte_ip.h b/lib/net/rte_ip.h
> >> index 05948b69b7..fd08ea31b2 100644
> >> --- a/lib/net/rte_ip.h
> >> +++ b/lib/net/rte_ip.h
> >> @@ -488,7 +488,7 @@ rte_ipv6_phdr_cksum(const struct rte_ipv6_hdr
> >> *ipv6_hdr, uint64_t ol_flags)
> >> }
> >>
> >> /**
> >> - * @internal Calculate the non-complemented IPv4 L4 checksum
> >> + * @internal Calculate the non-complemented IPv6 L4 checksum
> >> */
> >> static inline uint16_t
> >> __rte_ipv6_udptcp_cksum(const struct rte_ipv6_hdr *ipv6_hdr, const
> >> void *l4_hdr)
> >> @@ -509,15 +509,15 @@ __rte_ipv6_udptcp_cksum(const struct
> rte_ipv6_hdr
> >> *ipv6_hdr, const void *l4_hdr)
> >> /**
> >> * Process the IPv6 UDP or TCP checksum.
> >> *
> >> - * The IPv4 header should not contains options. The layer 4
> checksum
> >> - * must be set to 0 in the packet by the caller.
> >> + * The IPv6 header must not be followed by extension headers. The
> >> layer 4
> >> + * checksum must be set to 0 in the L4 header by the caller.
> >> *
> >> * @param ipv6_hdr
> >> * The pointer to the contiguous IPv6 header.
> >> * @param l4_hdr
> >> * The pointer to the beginning of the L4 header.
> >> * @return
> >> - * The complemented checksum to set in the IP packet.
> >> + * The complemented checksum to set in the L4 header.
>
> Isn't this wrong for 'rte_ipv4_udptcp_cksum()' too? Since you are
> touching this,
> can you fix that one too?
>
Ferruh, the description of the return value is technically correct; it mentions
the IP packet, not the IP header.
So this change is a clarification only.
However, I agree that the same clarification would also benefit
'rte_ipv4_udptcp_cksum()'.
> >> */
> >> static inline uint16_t
> >> rte_ipv6_udptcp_cksum(const struct rte_ipv6_hdr *ipv6_hdr, const
> void
> >> *l4_hdr)
> >> --
> >> 2.25.1
> >
> > Reviewed-by: Morten Brørup
> >
>
Re: [dpdk-dev] [PATCH] Enable AddressSanitizer feature on DPDK
On Tue, Sep 21, 2021 at 1:59 PM David Marchand wrote: > > On Mon, Sep 20, 2021 at 9:41 PM David Christensen > wrote: > > >>> We do not have a ppc platform, so there is no adaptation. > > >>> doc/guides/prog_guide/asan.rst has stated that we currently only > > >>> support Linux x86_64. You can adapt according to the following > > >>> documents, > > >> the main work is to modify the base address according to the platform. > > >>> Documents: > > >>> https://github.com/google/sanitizers/wiki/AddressSanitizer > > >>> https://github.com/llvm/llvm-project/tree/main/compiler-rt > > >> > > >> Understand you don't have such a platform. I looked into it and suggest > > >> the > > >> following change in lib/eal/common/malloc_elem.h: > > >> > > >> #define ASAN_SHADOW_GRAIN_SIZE 8 > > >> #define ASAN_SHADOW_SCALE 3 > > >> #ifdef RTE_ARCH_PPC_64 > > >> #define ASAN_SHADOW_OFFSET 0x0200 #else #define > > >> ASAN_SHADOW_OFFSET 0x7fff8000 #endif > > >> #define ASAN_MEM_FREE_FLAG 0xfd > > >> #define ASAN_MEM_REDZONE_FLAG 0xfa > > >> #define ASAN_MEM_TO_SHADOW(mem) (((mem) >> > > >> ASAN_SHADOW_SCALE) + > > >> ASAN_SHADOW_OFFSET) > > >> > > >> > > >> This resolves the segmentation error I receive. > > >> > > >> Dave > > >> > > > > > > Great, good information for dpdk asan tool. Because we can't do many > > > tests, > > > so when this patch is merged into the main line, you can submit the ppc > > > architecture patch. > > > > If your argument is that this is x86 only then please ensure it can't be > > enabled on non-x86 platforms such as ARM and PPC. I can then easily > > submit a follow-on patch to enable for PPC. > > > > As the patch currently stands it enables ASAN on a non-tested platform > > and provides an unexpected error for some users when it can easily be > > avoided. I'd advise not accepting the patch as currently presented. > > Please make sure only x86_64 gets this code enabled. I think, we need to opt out only the PPC. According https://developer.android.com/ndk/guides/asan, Arm64 and x86 are supported. I suggest resping by opting out PPC, I can test the next version and confirm the behavior on arm64 and give Ack. > I'll wait for a new revision, thanks. > > > -- > David Marchand >
[dpdk-dev] [v8, 0/4] cnxk: enable telemetry endpoints for mempool and ethdev
This patch series enables telemetry for cnxk in the following:
- NPA LF
- Mempool driver
- NIX LF
- Ethdev driver
Depends-on: series-18612 ("net/cnxk: support for inline ipsec")
v8:
- removed lib telemetry and crypto patches for seperate series.
Gowrishankar Muthukrishnan (4):
common/cnxk: add telemetry endpoints to npa
mempool/cnxk: add telemetry end points
common/cnxk: add telemetry endpoints to nix
net/cnxk: add telemetry endpoing to ethdev
drivers/common/cnxk/cnxk_telemetry.h | 26 +
drivers/common/cnxk/cnxk_telemetry_nix.c | 849 ++
drivers/common/cnxk/cnxk_telemetry_npa.c | 224 +
drivers/common/cnxk/meson.build | 7 +-
drivers/common/cnxk/roc_nix.c | 3 +
drivers/common/cnxk/roc_nix_priv.h| 9 +
drivers/common/cnxk/roc_nix_queue.c | 15 +-
drivers/common/cnxk/roc_platform.h| 15 +
drivers/mempool/cnxk/cnxk_mempool_telemetry.c | 100 +++
drivers/mempool/cnxk/meson.build | 1 +
drivers/net/cnxk/cnxk_ethdev_telemetry.c | 129 +++
drivers/net/cnxk/meson.build | 1 +
12 files changed, 1374 insertions(+), 5 deletions(-)
create mode 100644 drivers/common/cnxk/cnxk_telemetry.h
create mode 100644 drivers/common/cnxk/cnxk_telemetry_nix.c
create mode 100644 drivers/common/cnxk/cnxk_telemetry_npa.c
create mode 100644 drivers/mempool/cnxk/cnxk_mempool_telemetry.c
create mode 100644 drivers/net/cnxk/cnxk_ethdev_telemetry.c
--
2.25.1
[dpdk-dev] [v8, 1/4] common/cnxk: add telemetry endpoints to npa
Add telemetry endpoints to npa.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/common/cnxk/cnxk_telemetry.h | 26 +++
drivers/common/cnxk/cnxk_telemetry_npa.c | 224 +++
drivers/common/cnxk/meson.build | 6 +-
drivers/common/cnxk/roc_platform.h | 12 ++
4 files changed, 266 insertions(+), 2 deletions(-)
create mode 100644 drivers/common/cnxk/cnxk_telemetry.h
create mode 100644 drivers/common/cnxk/cnxk_telemetry_npa.c
diff --git a/drivers/common/cnxk/cnxk_telemetry.h
b/drivers/common/cnxk/cnxk_telemetry.h
new file mode 100644
index 00..1461fd893f
--- /dev/null
+++ b/drivers/common/cnxk/cnxk_telemetry.h
@@ -0,0 +1,26 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Marvell.
+ */
+
+#ifndef __CNXK_TELEMETRY_H_
+#define __CNXK_TELEMETRY_H_
+
+#define CNXK_TEL_STR(s) #s
+#define CNXK_TEL_STR_PREFIX(s, p) CNXK_TEL_STR(p##s)
+#define CNXK_TEL_DICT_INT(d, p, s, ...)
\
+ plt_tel_data_add_dict_int(d, CNXK_TEL_STR_PREFIX(s, __VA_ARGS__), \
+ (p)->s)
+#define CNXK_TEL_DICT_PTR(d, p, s, ...)
\
+ plt_tel_data_add_dict_ptr(d, CNXK_TEL_STR_PREFIX(s, __VA_ARGS__), \
+ (void *)(p)->s)
+#define CNXK_TEL_DICT_BF_PTR(d, p, s, ...)
\
+ plt_tel_data_add_dict_ptr(d, CNXK_TEL_STR_PREFIX(s, __VA_ARGS__), \
+ (void *)(uint64_t)(p)->s)
+#define CNXK_TEL_DICT_U64(d, p, s, ...)
\
+ plt_tel_data_add_dict_u64(d, CNXK_TEL_STR_PREFIX(s, __VA_ARGS__), \
+ (p)->s)
+#define CNXK_TEL_DICT_STR(d, p, s, ...)
\
+ plt_tel_data_add_dict_string(d, CNXK_TEL_STR_PREFIX(s, __VA_ARGS__), \
+(p)->s)
+
+#endif /* __CNXK_TELEMETRY_H_ */
diff --git a/drivers/common/cnxk/cnxk_telemetry_npa.c
b/drivers/common/cnxk/cnxk_telemetry_npa.c
new file mode 100644
index 00..37c5f2e961
--- /dev/null
+++ b/drivers/common/cnxk/cnxk_telemetry_npa.c
@@ -0,0 +1,224 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell.
+ */
+
+#include "cnxk_telemetry.h"
+#include "roc_api.h"
+#include "roc_priv.h"
+
+static int
+cnxk_tel_npa(struct plt_tel_data *d)
+{
+ struct npa_lf *lf;
+ int aura_cnt = 0;
+ uint32_t i;
+
+ lf = idev_npa_obj_get();
+ if (lf == NULL)
+ return NPA_ERR_DEVICE_NOT_BOUNDED;
+
+ for (i = 0; i < lf->nr_pools; i++) {
+ if (plt_bitmap_get(lf->npa_bmp, i))
+ continue;
+ aura_cnt++;
+ }
+
+ plt_tel_data_add_dict_ptr(d, "npa", lf);
+ plt_tel_data_add_dict_int(d, "pf", dev_get_pf(lf->pf_func));
+ plt_tel_data_add_dict_int(d, "vf", dev_get_vf(lf->pf_func));
+ plt_tel_data_add_dict_int(d, "aura_cnt", aura_cnt);
+
+ CNXK_TEL_DICT_PTR(d, lf, pci_dev);
+ CNXK_TEL_DICT_PTR(d, lf, npa_bmp);
+ CNXK_TEL_DICT_PTR(d, lf, npa_bmp_mem);
+ CNXK_TEL_DICT_PTR(d, lf, npa_qint_mem);
+ CNXK_TEL_DICT_PTR(d, lf, intr_handle);
+ CNXK_TEL_DICT_PTR(d, lf, mbox);
+ CNXK_TEL_DICT_PTR(d, lf, base);
+ CNXK_TEL_DICT_INT(d, lf, stack_pg_ptrs);
+ CNXK_TEL_DICT_INT(d, lf, stack_pg_bytes);
+ CNXK_TEL_DICT_INT(d, lf, npa_msixoff);
+ CNXK_TEL_DICT_INT(d, lf, nr_pools);
+ CNXK_TEL_DICT_INT(d, lf, pf_func);
+ CNXK_TEL_DICT_INT(d, lf, aura_sz);
+ CNXK_TEL_DICT_INT(d, lf, qints);
+
+ return 0;
+}
+
+static int
+cnxk_tel_npa_aura(int aura_id, struct plt_tel_data *d)
+{
+ __io struct npa_aura_s *aura;
+ struct npa_aq_enq_req *req;
+ struct npa_aq_enq_rsp *rsp;
+ struct npa_lf *lf;
+ int rc;
+
+ lf = idev_npa_obj_get();
+ if (lf == NULL)
+ return NPA_ERR_DEVICE_NOT_BOUNDED;
+
+ if (plt_bitmap_get(lf->npa_bmp, aura_id))
+ return -1;
+
+ req = mbox_alloc_msg_npa_aq_enq(lf->mbox);
+ if (!req) {
+ plt_err("Failed to alloc aq enq for npa");
+ return -1;
+ }
+
+ req->aura_id = aura_id;
+ req->ctype = NPA_AQ_CTYPE_AURA;
+ req->op = NPA_AQ_INSTOP_READ;
+
+ rc = mbox_process_msg(lf->mbox, (void *)&rsp);
+ if (rc) {
+ plt_err("Failed to get pool(%d) context", aura_id);
+ return rc;
+ }
+
+ aura = &rsp->aura;
+ CNXK_TEL_DICT_PTR(d, aura, pool_addr, w0_);
+ CNXK_TEL_DICT_INT(d, aura, ena, w1_);
+ CNXK_TEL_DICT_INT(d, aura, pool_caching, w1_);
+ CNXK_TEL_DICT_INT(d, aura, pool_way_mask, w1_);
+ CNXK_TEL_DICT_INT(d, aura, avg_con, w1_);
+ CNXK_TEL_DICT_INT(d, aura, pool_drop_ena, w1_);
+ CNXK_TEL_DICT_INT(d, aura, aura_drop_ena, w1_);
+ C
[dpdk-dev] [v8, 2/4] mempool/cnxk: add telemetry end points
Adding telemetry end points to cnxk mempool driver.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/mempool/cnxk/cnxk_mempool_telemetry.c | 100 ++
drivers/mempool/cnxk/meson.build | 1 +
2 files changed, 101 insertions(+)
create mode 100644 drivers/mempool/cnxk/cnxk_mempool_telemetry.c
diff --git a/drivers/mempool/cnxk/cnxk_mempool_telemetry.c
b/drivers/mempool/cnxk/cnxk_mempool_telemetry.c
new file mode 100644
index 00..690f557f62
--- /dev/null
+++ b/drivers/mempool/cnxk/cnxk_mempool_telemetry.c
@@ -0,0 +1,100 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell.
+ */
+
+#include
+#include
+#include
+
+#include
+
+#include "cnxk_mempool.h"
+#include "cnxk_telemetry.h"
+
+static void
+mempool_list_cb(struct rte_mempool *mp, void *arg)
+{
+ struct rte_tel_data *d = (struct rte_tel_data *)arg;
+
+ rte_tel_data_add_array_string(d, mp->name);
+}
+
+static int
+mempool_tel_handle_list(const char *cmd __rte_unused,
+ const char *params __rte_unused, struct rte_tel_data *d)
+{
+ rte_tel_data_start_array(d, RTE_TEL_STRING_VAL);
+ rte_mempool_walk(mempool_list_cb, d);
+ return 0;
+}
+
+struct mempool_info_cb_arg {
+ char *pool_name;
+ struct rte_tel_data *d;
+};
+
+static void
+mempool_info_cb(struct rte_mempool *mp, void *arg)
+{
+ struct mempool_info_cb_arg *info = (struct mempool_info_cb_arg *)arg;
+ const struct rte_memzone *mz;
+ int aura_id;
+
+ if (strncmp(mp->name, info->pool_name, RTE_MEMZONE_NAMESIZE))
+ return;
+
+ CNXK_TEL_DICT_STR(info->d, mp, name);
+ CNXK_TEL_DICT_INT(info->d, mp, pool_id);
+ CNXK_TEL_DICT_INT(info->d, mp, flags);
+ CNXK_TEL_DICT_INT(info->d, mp, socket_id);
+ CNXK_TEL_DICT_INT(info->d, mp, size);
+ CNXK_TEL_DICT_INT(info->d, mp, cache_size);
+ CNXK_TEL_DICT_INT(info->d, mp, elt_size);
+ CNXK_TEL_DICT_INT(info->d, mp, header_size);
+ CNXK_TEL_DICT_INT(info->d, mp, trailer_size);
+ CNXK_TEL_DICT_INT(info->d, mp, private_data_size);
+ CNXK_TEL_DICT_INT(info->d, mp, ops_index);
+ CNXK_TEL_DICT_INT(info->d, mp, populated_size);
+
+ aura_id = roc_npa_aura_handle_to_aura(mp->pool_id);
+ rte_tel_data_add_dict_int(info->d, "aura_id", aura_id);
+
+ mz = mp->mz;
+ CNXK_TEL_DICT_STR(info->d, mz, name, mz_);
+ CNXK_TEL_DICT_PTR(info->d, mz, iova, mz_);
+ CNXK_TEL_DICT_PTR(info->d, mz, addr, mz_);
+ CNXK_TEL_DICT_INT(info->d, mz, len, mz_);
+ CNXK_TEL_DICT_U64(info->d, mz, hugepage_sz, mz_);
+ CNXK_TEL_DICT_INT(info->d, mz, socket_id, mz_);
+ CNXK_TEL_DICT_INT(info->d, mz, flags, mz_);
+}
+
+static int
+mempool_tel_handle_info(const char *cmd __rte_unused, const char *params,
+ struct rte_tel_data *d)
+{
+ struct mempool_info_cb_arg mp_arg;
+ char name[RTE_MEMZONE_NAMESIZE];
+
+ if (params == NULL || strlen(params) == 0)
+ return -1;
+
+ rte_strlcpy(name, params, RTE_MEMZONE_NAMESIZE);
+
+ rte_tel_data_start_dict(d);
+ mp_arg.pool_name = name;
+ mp_arg.d = d;
+ rte_mempool_walk(mempool_info_cb, &mp_arg);
+
+ return 0;
+}
+
+RTE_INIT(cnxk_mempool_init_telemetry)
+{
+ rte_telemetry_register_cmd(
+ "/cnxk/mempool/list", mempool_tel_handle_list,
+ "Returns list of available mempools. Takes no parameters");
+ rte_telemetry_register_cmd(
+ "/cnxk/mempool/info", mempool_tel_handle_info,
+ "Returns mempool info. Parameters: pool_name");
+}
diff --git a/drivers/mempool/cnxk/meson.build b/drivers/mempool/cnxk/meson.build
index e28a9e044d..d5d1978569 100644
--- a/drivers/mempool/cnxk/meson.build
+++ b/drivers/mempool/cnxk/meson.build
@@ -11,6 +11,7 @@ endif
sources = files(
'cnxk_mempool.c',
'cnxk_mempool_ops.c',
+'cnxk_mempool_telemetry.c',
'cn9k_mempool_ops.c',
'cn10k_mempool_ops.c',
)
--
2.25.1
[dpdk-dev] [v8, 3/4] common/cnxk: add telemetry endpoints to nix
Add telemetry endpoints to nix.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/common/cnxk/cnxk_telemetry_nix.c | 849 +++
drivers/common/cnxk/meson.build | 3 +-
drivers/common/cnxk/roc_nix.c| 3 +
drivers/common/cnxk/roc_nix_priv.h | 9 +
drivers/common/cnxk/roc_nix_queue.c | 15 +-
drivers/common/cnxk/roc_platform.h | 3 +
6 files changed, 878 insertions(+), 4 deletions(-)
create mode 100644 drivers/common/cnxk/cnxk_telemetry_nix.c
diff --git a/drivers/common/cnxk/cnxk_telemetry_nix.c
b/drivers/common/cnxk/cnxk_telemetry_nix.c
new file mode 100644
index 00..1175f68a51
--- /dev/null
+++ b/drivers/common/cnxk/cnxk_telemetry_nix.c
@@ -0,0 +1,849 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell.
+ */
+
+#include "cnxk_telemetry.h"
+#include "roc_api.h"
+#include "roc_priv.h"
+
+struct nix_tel_node {
+ TAILQ_ENTRY(nix_tel_node) node;
+ struct roc_nix *nix;
+ uint16_t n_rq;
+ uint16_t n_cq;
+ uint16_t n_sq;
+ struct roc_nix_rq **rqs;
+ struct roc_nix_cq **cqs;
+ struct roc_nix_sq **sqs;
+};
+
+TAILQ_HEAD(nix_tel_node_list, nix_tel_node);
+static struct nix_tel_node_list nix_list;
+
+static struct nix_tel_node *
+nix_tel_node_get(struct roc_nix *roc_nix)
+{
+ struct nix_tel_node *node, *roc_node = NULL;
+
+ TAILQ_FOREACH(node, &nix_list, node) {
+ if (node->nix == roc_nix) {
+ roc_node = node;
+ break;
+ }
+ }
+
+ return roc_node;
+}
+
+int
+nix_tel_node_add(struct roc_nix *roc_nix)
+{
+ struct nix *nix = roc_nix_to_nix_priv(roc_nix);
+ struct nix_tel_node *node;
+
+ node = nix_tel_node_get(roc_nix);
+ if (node) {
+ if (nix->nb_rx_queues == node->n_rq &&
+ nix->nb_tx_queues == node->n_sq)
+ return 0;
+
+ nix_tel_node_del(roc_nix);
+ }
+
+ node = plt_zmalloc(sizeof(struct nix_tel_node), 0);
+ if (!node)
+ return -1;
+
+ node->nix = roc_nix;
+ node->rqs =
+ plt_zmalloc(nix->nb_rx_queues * sizeof(struct roc_nix_rq *), 0);
+ node->cqs =
+ plt_zmalloc(nix->nb_rx_queues * sizeof(struct roc_nix_cq *), 0);
+ node->sqs =
+ plt_zmalloc(nix->nb_tx_queues * sizeof(struct roc_nix_sq *), 0);
+ TAILQ_INSERT_TAIL(&nix_list, node, node);
+
+ return 0;
+}
+
+void
+nix_tel_node_del(struct roc_nix *roc_nix)
+{
+ struct nix_tel_node *node;
+
+ TAILQ_FOREACH(node, &nix_list, node) {
+ if (node->nix == roc_nix) {
+ plt_free(node->rqs);
+ plt_free(node->cqs);
+ plt_free(node->sqs);
+ TAILQ_REMOVE(&nix_list, node, node);
+ }
+ }
+
+ plt_free(node);
+}
+
+static struct nix_tel_node *
+nix_tel_node_get_by_pcidev_name(const char *name)
+{
+ struct nix_tel_node *node, *roc_node = NULL;
+
+ TAILQ_FOREACH(node, &nix_list, node) {
+ if (!strncmp(node->nix->pci_dev->name, name,
+PCI_PRI_STR_SIZE)) {
+ roc_node = node;
+ break;
+ }
+ }
+
+ return roc_node;
+}
+
+int
+nix_tel_node_add_rq(struct roc_nix_rq *rq)
+{
+ struct nix_tel_node *node;
+
+ node = nix_tel_node_get(rq->roc_nix);
+ if (!node)
+ return -1;
+
+ node->rqs[rq->qid] = rq;
+ node->n_rq++;
+ return 0;
+}
+
+int
+nix_tel_node_add_cq(struct roc_nix_cq *cq)
+{
+ struct nix_tel_node *node;
+
+ node = nix_tel_node_get(cq->roc_nix);
+ if (!node)
+ return -1;
+
+ node->cqs[cq->qid] = cq;
+ node->n_cq++;
+ return 0;
+}
+
+int
+nix_tel_node_add_sq(struct roc_nix_sq *sq)
+{
+ struct nix_tel_node *node;
+
+ node = nix_tel_node_get(sq->roc_nix);
+ if (!node)
+ return -1;
+
+ node->sqs[sq->qid] = sq;
+ node->n_sq++;
+ return 0;
+}
+
+static int
+cnxk_tel_nix(struct roc_nix *roc_nix, struct plt_tel_data *d)
+{
+ struct nix *nix = roc_nix_to_nix_priv(roc_nix);
+
+ struct dev *dev = &nix->dev;
+
+ plt_tel_data_add_dict_ptr(d, "nix", nix);
+ plt_tel_data_add_dict_int(d, "pf_func", dev->pf_func);
+ plt_tel_data_add_dict_int(d, "pf", dev_get_pf(dev->pf_func));
+ plt_tel_data_add_dict_int(d, "vf", dev_get_vf(dev->pf_func));
+
+ CNXK_TEL_DICT_PTR(d, dev, bar2);
+ CNXK_TEL_DICT_PTR(d, dev, bar4);
+ CNXK_TEL_DICT_INT(d, roc_nix, port_id);
+ CNXK_TEL_DICT_INT(d, roc_nix, rss_tag_as_xor);
+ CNXK_TEL_DICT_INT(d, roc_nix, max_sqb_count);
+ CNXK_TEL_DICT_PTR(d, nix, pci_dev);
+ CNXK_TEL_DICT_PTR(d, nix, base);
+ CNXK_TEL_DICT_PTR(d, nix, lmt_base);
+ CNXK_TEL_DICT
[dpdk-dev] [v8, 4/4] net/cnxk: add telemetry endpoing to ethdev
Add telemetry endpoint to ethdev.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/net/cnxk/cnxk_ethdev_telemetry.c | 129 +++
drivers/net/cnxk/meson.build | 1 +
2 files changed, 130 insertions(+)
create mode 100644 drivers/net/cnxk/cnxk_ethdev_telemetry.c
diff --git a/drivers/net/cnxk/cnxk_ethdev_telemetry.c
b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
new file mode 100644
index 00..de8c468670
--- /dev/null
+++ b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
@@ -0,0 +1,129 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell International Ltd.
+ */
+
+#include
+
+#include "cnxk_ethdev.h"
+
+/* Macro to count no of words in eth_info_s size */
+#define ETH_INFO_SZ
\
+ (RTE_ALIGN_CEIL(sizeof(struct eth_info_s), sizeof(uint64_t)) / \
+sizeof(uint64_t))
+#define MACADDR_LEN 18
+
+static int
+ethdev_tel_handle_info(const char *cmd __rte_unused,
+ const char *params __rte_unused, struct rte_tel_data *d)
+{
+ struct rte_eth_dev *eth_dev;
+ struct rte_tel_data *i_data;
+ struct cnxk_eth_dev *dev;
+ union eth_info_u {
+ struct eth_info_s {
+ /** PF/VF information */
+ uint16_t pf_func;
+ /** No of rx queues */
+ uint16_t rx_queues;
+ /** No of tx queues */
+ uint16_t tx_queues;
+ /** Port ID */
+ uint16_t port_id;
+ /** MAC entries */
+ char mac_addr[MACADDR_LEN];
+ uint8_t max_mac_entries;
+ bool dmac_filter_ena;
+ uint8_t dmac_filter_count;
+ uint16_t flags;
+ uint8_t ptype_disable;
+ bool scalar_ena;
+ bool ptp_ena;
+ /* Offload capabilities */
+ uint64_t rx_offload_capa;
+ uint64_t tx_offload_capa;
+ uint32_t speed_capa;
+ /* Configured offloads */
+ uint64_t rx_offloads;
+ uint64_t tx_offloads;
+ /* Platform specific offload flags */
+ uint16_t rx_offload_flags;
+ uint16_t tx_offload_flags;
+ /* ETHDEV RSS HF bitmask */
+ uint64_t ethdev_rss_hf;
+ } info;
+ uint64_t val[ETH_INFO_SZ];
+ } eth_info;
+ struct eth_info_s *info;
+ unsigned int i, j = 0;
+ int n_ports;
+
+ n_ports = rte_eth_dev_count_avail();
+ if (!n_ports) {
+ plt_err("No active ethernet ports found.");
+ return -1;
+ }
+
+ rte_tel_data_start_dict(d);
+ rte_tel_data_add_dict_int(d, "n_ports", n_ports);
+
+ i_data = rte_tel_data_alloc();
+ rte_tel_data_start_array(i_data, RTE_TEL_U64_VAL);
+
+ for (i = 0; i < RTE_MAX_ETHPORTS; i++) {
+ /* Skip if port is unused */
+ if (!rte_eth_dev_is_valid_port(i))
+ continue;
+
+ eth_dev = &rte_eth_devices[i];
+ if (eth_dev) {
+ memset(ð_info, 0, sizeof(eth_info));
+ info = ð_info.info;
+ dev = cnxk_eth_pmd_priv(eth_dev);
+ if (dev) {
+ info->pf_func = roc_nix_get_pf_func(&dev->nix);
+ memset(info->mac_addr, 0, MACADDR_LEN);
+ snprintf(info->mac_addr, MACADDR_LEN,
+"%02x:%02x:%02x:%02x:%02x:%02x",
+dev->mac_addr[0], dev->mac_addr[1],
+dev->mac_addr[2], dev->mac_addr[3],
+dev->mac_addr[4], dev->mac_addr[5]);
+ info->max_mac_entries = dev->max_mac_entries;
+ info->dmac_filter_ena = dev->dmac_filter_enable;
+ info->dmac_filter_count =
+ dev->dmac_filter_count;
+ info->flags = dev->flags;
+ info->ptype_disable = dev->ptype_disable;
+ info->scalar_ena = dev->scalar_ena;
+ info->ptp_ena = dev->ptp_en;
+ info->rx_offload_capa = dev->rx_offload_capa;
+ info->tx_offload_capa = dev->tx_offload_capa;
+ info->rx_offloads = dev->rx_offloads;
+ info->tx_offloads = dev->tx_offloads;
+
[dpdk-dev] [PATCH 1/2] common/cnxk: clear rvum interrupts
As per an known HW issue RVUM interrupts may get dropped, If an RVUM
interrupt event occurs when PCCPF_XXX_MSIX_CAP_HDR[MSIXEN]=0 then no
interrupt is triggered, which is expected. But after MSIXEN is set to
1, subsequently if same interrupts event occurs again, still no
interrupt will be triggered.
As a workaround, all RVUM interrupt lines should be cleared between
MSIXEN=0 and MSIXEN=1.
Signed-off-by: Harman Kalra
---
drivers/common/cnxk/roc_dev.c | 35 +++
1 file changed, 35 insertions(+)
diff --git a/drivers/common/cnxk/roc_dev.c b/drivers/common/cnxk/roc_dev.c
index 4e204373dc..ce6980cbe4 100644
--- a/drivers/common/cnxk/roc_dev.c
+++ b/drivers/common/cnxk/roc_dev.c
@@ -884,6 +884,38 @@ vf_flr_register_irqs(struct plt_pci_device *pci_dev,
struct dev *dev)
return 0;
}
+static void
+clear_rvum_interrupts(struct dev *dev)
+{
+ uint64_t intr;
+ int i;
+
+ if (dev_is_vf(dev)) {
+ /* Clear VF mbox interrupt */
+ intr = plt_read64(dev->bar2 + RVU_VF_INT);
+ if (intr)
+ plt_write64(intr, dev->bar2 + RVU_VF_INT);
+ } else {
+ /* Clear AF PF interrupt line */
+ intr = plt_read64(dev->bar2 + RVU_PF_INT);
+ if (intr)
+ plt_write64(intr, dev->bar2 + RVU_PF_INT);
+ for (i = 0; i < MAX_VFPF_DWORD_BITS; ++i) {
+ /* Clear MBOX interrupts */
+ intr = plt_read64(dev->bar2 + RVU_PF_VFPF_MBOX_INTX(i));
+ if (intr)
+ plt_write64(intr,
+ dev->bar2 +
+ RVU_PF_VFPF_MBOX_INTX(i));
+ /* Clear VF FLR interrupts */
+ intr = plt_read64(dev->bar2 + RVU_PF_VFFLR_INTX(i));
+ if (intr)
+ plt_write64(intr,
+ dev->bar2 + RVU_PF_VFFLR_INTX(i));
+ }
+ }
+}
+
int
dev_active_vfs(struct dev *dev)
{
@@ -1090,6 +1122,9 @@ dev_init(struct dev *dev, struct plt_pci_device *pci_dev)
intr_offset = RVU_PF_INT;
}
+ /* Clear all RVUM interrupts */
+ clear_rvum_interrupts(dev);
+
/* Initialize the local mbox */
rc = mbox_init(&dev->mbox_local, mbox, bar2, direction, 1, intr_offset);
if (rc)
--
2.18.0
[dpdk-dev] [PATCH 2/2] common/cnxk: cq overflow issue
An issue exists on some HW revisions whereby if a CQ overflows
NIX may have undefined behavior, e.g. free incorrect buffers.
Implementing a workaround for this known HW issue.
Signed-off-by: Harman Kalra
---
drivers/common/cnxk/roc_nix_priv.h | 3 ++-
drivers/common/cnxk/roc_nix_queue.c | 18 +++---
2 files changed, 17 insertions(+), 4 deletions(-)
diff --git a/drivers/common/cnxk/roc_nix_priv.h
b/drivers/common/cnxk/roc_nix_priv.h
index 9dc0c88a6f..1bd1b6a36b 100644
--- a/drivers/common/cnxk/roc_nix_priv.h
+++ b/drivers/common/cnxk/roc_nix_priv.h
@@ -17,7 +17,8 @@
/* Apply BP/DROP when CQ is 95% full */
#define NIX_CQ_THRESH_LEVEL(5 * 256 / 100)
-#define NIX_RQ_AURA_THRESH(x) (((x) * 95) / 100)
+#define NIX_CQ_FULL_ERRATA_SKID (1024ull * 256)
+#define NIX_RQ_AURA_THRESH(x) (((x)*95) / 100)
/* IRQ triggered when NIX_LF_CINTX_CNT[QCOUNT] crosses this value */
#define CQ_CQE_THRESH_DEFAULT 0x1ULL
diff --git a/drivers/common/cnxk/roc_nix_queue.c
b/drivers/common/cnxk/roc_nix_queue.c
index 76e439e7a9..d7c4844d69 100644
--- a/drivers/common/cnxk/roc_nix_queue.c
+++ b/drivers/common/cnxk/roc_nix_queue.c
@@ -2,6 +2,8 @@
* Copyright(C) 2021 Marvell.
*/
+#include
+
#include "roc_api.h"
#include "roc_priv.h"
@@ -435,7 +437,6 @@ roc_nix_cq_init(struct roc_nix *roc_nix, struct roc_nix_cq
*cq)
cq->status = (int64_t *)(nix->base + NIX_LF_CQ_OP_STATUS);
cq->wdata = (uint64_t)cq->qid << 32;
cq->roc_nix = roc_nix;
- cq->drop_thresh = NIX_CQ_THRESH_LEVEL;
/* CQE of W16 */
desc_sz = cq->nb_desc * NIX_CQ_ENTRY_SZ;
@@ -476,8 +477,19 @@ roc_nix_cq_init(struct roc_nix *roc_nix, struct roc_nix_cq
*cq)
/* Map CQ0 [RQ0] to CINT0 and so on till max 64 irqs */
cq_ctx->cint_idx = cq->qid;
- cq_ctx->drop = cq->drop_thresh;
- cq_ctx->drop_ena = 1;
+ if (roc_model_is_cn96_a0() || roc_model_is_cn95_a0()) {
+ const float rx_cq_skid = NIX_CQ_FULL_ERRATA_SKID;
+ uint16_t min_rx_drop;
+
+ min_rx_drop = ceil(rx_cq_skid / (float)cq->nb_desc);
+ cq_ctx->drop = min_rx_drop;
+ cq_ctx->drop_ena = 1;
+ cq->drop_thresh = min_rx_drop;
+ } else {
+ cq->drop_thresh = NIX_CQ_THRESH_LEVEL;
+ cq_ctx->drop = cq->drop_thresh;
+ cq_ctx->drop_ena = 1;
+ }
/* TX pause frames enable flow ctrl on RX side */
if (nix->tx_pause) {
--
2.18.0
[dpdk-dev] [v2] telemetry: fix json output buffer size
Fix json output buffer size for a single largest value.
Fixes: 52af6ccb2b39 ("telemetry: add utility functions for creating JSON")
v2:
- split from series 18768 ("cnxk: enable telemetry endpoints").
Signed-off-by: Gowrishankar Muthukrishnan
---
lib/telemetry/telemetry_json.h | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/lib/telemetry/telemetry_json.h b/lib/telemetry/telemetry_json.h
index ad270b9b30..ba2fde34cb 100644
--- a/lib/telemetry/telemetry_json.h
+++ b/lib/telemetry/telemetry_json.h
@@ -9,6 +9,7 @@
#include
#include
#include
+#include
/**
* @file
@@ -23,13 +24,15 @@
* @internal
* Copies a value into a buffer if the buffer has enough available space.
* Nothing written to buffer if an overflow ocurs.
- * This function is not for use for values larger than 1k.
+ * Size of buffer is (single largest value - 6), where at least 6 chars
+ * would have been used for creating json dict i.e '{"x": ... }'.
+ * This function is not for use for values larger than this buffer size.
*/
__rte_format_printf(3, 4)
static inline int
__json_snprintf(char *buf, const int len, const char *format, ...)
{
- char tmp[1024];
+ char tmp[RTE_TEL_MAX_SINGLE_STRING_LEN - 6];
va_list ap;
int ret;
--
2.25.1
Re: [dpdk-dev] |FAILURE| pw97277 [PATCH] [v4] eal: read data buffer on RTE_INTR_HANDLE_VFIO_REQ
Anatoly, Jeff, I would appreciate review. Thanks, Maciej -Original Message- From: David Marchand Sent: Friday, September 3, 2021 2:01 PM To: Szwed, Maciej ; Burakov, Anatoly ; Jeff Guo Cc: dev@dpdk.org Subject: Re: [dpdk-dev] |FAILURE| pw97277 [PATCH] [v4] eal: read data buffer on RTE_INTR_HANDLE_VFIO_REQ On Fri, Sep 3, 2021 at 10:59 AM Szwed, Maciej wrote: > -Original Message- > From: dev On Behalf Of Szwed, Maciej > Sent: Monday, August 30, 2021 11:34 AM > To: dev@dpdk.org > Subject: Re: [dpdk-dev] |FAILURE| pw97277 [PATCH] [v4] eal: read data > buffer on RTE_INTR_HANDLE_VFIO_REQ > > Hi, > I'm new to posting patches to DPDK repository. I've submitted a patch > recently and one of the tests performed on it failed. I went through the > failed test log and it looks like it does not have anything to do with my > patch. You can check details below. What should be the next steps I should > perform? > About this specific failure in the CI, this is a known issue. This failure here won't block acceptance of this patch. But we need a review/ack for this patch. When submitting a patch, you should copy maintainers and people who worked on this part of DPDK. Anatoly, Jeff, can you have a look? Thanks. -- David Marchand
Re: [dpdk-dev] [v8, 4/4] net/cnxk: add telemetry endpoing to ethdev
On Tue, Sep 21, 2021 at 4:23 PM Gowrishankar Muthukrishnan
wrote:
>
> Add telemetry endpoint to ethdev.
>
> Signed-off-by: Gowrishankar Muthukrishnan
> ---
> drivers/net/cnxk/cnxk_ethdev_telemetry.c | 129 +++
> drivers/net/cnxk/meson.build | 1 +
> 2 files changed, 130 insertions(+)
> create mode 100644 drivers/net/cnxk/cnxk_ethdev_telemetry.c
>
> diff --git a/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> new file mode 100644
> index 00..de8c468670
> --- /dev/null
> +++ b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> @@ -0,0 +1,129 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2021 Marvell International Ltd.
> + */
> +
> +#include
> +
> +#include "cnxk_ethdev.h"
> +
> +/* Macro to count no of words in eth_info_s size */
> +#define ETH_INFO_SZ
> \
> + (RTE_ALIGN_CEIL(sizeof(struct eth_info_s), sizeof(uint64_t)) /
> \
> +sizeof(uint64_t))
> +#define MACADDR_LEN 18
> +
> +static int
> +ethdev_tel_handle_info(const char *cmd __rte_unused,
> + const char *params __rte_unused, struct rte_tel_data
> *d)
> +{
> + struct rte_eth_dev *eth_dev;
> + struct rte_tel_data *i_data;
> + struct cnxk_eth_dev *dev;
> + union eth_info_u {
> + struct eth_info_s {
> + /** PF/VF information */
> + uint16_t pf_func;
> + /** No of rx queues */
> + uint16_t rx_queues;
> + /** No of tx queues */
> + uint16_t tx_queues;
> + /** Port ID */
> + uint16_t port_id;
> + /** MAC entries */
> + char mac_addr[MACADDR_LEN];
> + uint8_t max_mac_entries;
> + bool dmac_filter_ena;
> + uint8_t dmac_filter_count;
> + uint16_t flags;
> + uint8_t ptype_disable;
> + bool scalar_ena;
> + bool ptp_ena;
> + /* Offload capabilities */
> + uint64_t rx_offload_capa;
> + uint64_t tx_offload_capa;
> + uint32_t speed_capa;
> + /* Configured offloads */
> + uint64_t rx_offloads;
> + uint64_t tx_offloads;
> + /* Platform specific offload flags */
> + uint16_t rx_offload_flags;
> + uint16_t tx_offload_flags;
> + /* ETHDEV RSS HF bitmask */
> + uint64_t ethdev_rss_hf;
+ Ethdev, , mempool and telemetry maintainers
All of the above data except[1] is generic data that can be derived
from ethdev APIs or ethdev data from lib/ethdev itself,
IMO, We should avoid duplicating this in driver and make useful for
other driver/application by adding generic endpoint in ethdev.
Same comment for "[2/4] mempool/cnxk: add telemetry end points" for
mempool subsystem.
Thoughts from Maintainers?
[1]
uint8_t ptype_disable;
bool scalar_ena;
bool ptp_ena;
uint16_t flags;
[dpdk-dev] [v2] crypto/cnxk: add telemetry endpoints to cryptodev
Add telemetry endpoints to cryptodev.
Signed-off-by: Gowrishankar Muthukrishnan
---
Depends-on: series-19052 ("cnxk: enable telemetry endpoints for mempool and
ethdev")
Depends-on: patch-19054 ("telemetry: fix json output buffer size")
v2:
- split from series 18768 ("cnxk: enable telemetry endpoints").
---
.../crypto/cnxk/cnxk_cryptodev_telemetry.c| 154 ++
drivers/crypto/cnxk/meson.build | 1 +
2 files changed, 155 insertions(+)
create mode 100644 drivers/crypto/cnxk/cnxk_cryptodev_telemetry.c
diff --git a/drivers/crypto/cnxk/cnxk_cryptodev_telemetry.c
b/drivers/crypto/cnxk/cnxk_cryptodev_telemetry.c
new file mode 100644
index 00..4c754bc6c7
--- /dev/null
+++ b/drivers/crypto/cnxk/cnxk_cryptodev_telemetry.c
@@ -0,0 +1,154 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell.
+ */
+
+#include
+#include
+#include
+
+#include
+
+#include "cnxk_cryptodev.h"
+#include "cnxk_telemetry.h"
+
+#define CRYPTO_CAPS_SZ
\
+ (RTE_ALIGN_CEIL(sizeof(struct rte_cryptodev_capabilities), \
+ sizeof(uint64_t)) /\
+sizeof(uint64_t))
+
+#define SEC_CAPS_SZ
\
+ (RTE_ALIGN_CEIL(sizeof(struct rte_security_capability),\
+ sizeof(uint64_t)) /\
+sizeof(uint64_t))
+
+static int
+cryptodev_tel_handle_list(const char *cmd __rte_unused,
+ const char *params __rte_unused,
+ struct rte_tel_data *d)
+{
+ struct rte_cryptodev *cryptodev;
+ unsigned int i;
+
+ rte_tel_data_start_array(d, RTE_TEL_STRING_VAL);
+
+ for (i = 0; i < RTE_CRYPTO_MAX_DEVS; i++) {
+ cryptodev = rte_cryptodev_pmd_get_dev(i);
+ if (cryptodev->attached != RTE_CRYPTODEV_ATTACHED)
+ continue;
+
+ rte_tel_data_add_array_string(d, cryptodev->data->name);
+ }
+
+ return 0;
+}
+
+static int
+crypto_caps_array(struct rte_tel_data *d,
+ struct rte_cryptodev_capabilities *dev_caps,
+ size_t dev_caps_n)
+{
+ union caps_u {
+ struct rte_cryptodev_capabilities dev_caps;
+ uint64_t val[CRYPTO_CAPS_SZ];
+ } caps;
+ unsigned int i, j, n = 0;
+
+ rte_tel_data_start_array(d, RTE_TEL_U64_VAL);
+
+ for (i = 0; i < dev_caps_n; i++) {
+ if (dev_caps[i].op == RTE_CRYPTO_OP_TYPE_UNDEFINED)
+ break;
+
+ memset(&caps, 0, sizeof(caps));
+ rte_memcpy(&caps.dev_caps, &dev_caps[i], sizeof(dev_caps[0]));
+ for (j = 0; j < CRYPTO_CAPS_SZ; j++)
+ rte_tel_data_add_array_u64(d, caps.val[j]);
+ ++n;
+ }
+
+ return n;
+}
+
+static int
+sec_caps_array(struct rte_tel_data *d, struct rte_security_capability
*dev_caps,
+ size_t dev_caps_n)
+{
+ union caps_u {
+ struct rte_security_capability dev_caps;
+ uint64_t val[SEC_CAPS_SZ];
+ } caps;
+ unsigned int i, j, n = 0;
+
+ rte_tel_data_start_array(d, RTE_TEL_U64_VAL);
+
+ for (i = 0; i < dev_caps_n; i++) {
+ memset(&caps, 0, sizeof(caps));
+ rte_memcpy(&caps.dev_caps, &dev_caps[i], sizeof(dev_caps[0]));
+ for (j = 0; j < SEC_CAPS_SZ; j++)
+ rte_tel_data_add_array_u64(d, caps.val[j]);
+ ++n;
+ }
+
+ return n;
+}
+
+static int
+cryptodev_tel_handle_info(const char *cmd __rte_unused, const char *params,
+ struct rte_tel_data *d)
+{
+ struct rte_tel_data *crypto_caps, *sec_crypto_caps, *sec_caps;
+ int crypto_caps_n, sec_crypto_caps_n, sec_caps_n;
+ char name[RTE_CRYPTODEV_NAME_MAX_LEN];
+ struct rte_cryptodev *dev;
+ struct cnxk_cpt_vf *vf;
+
+ if (params == NULL || strlen(params) == 0 || !isdigit(*params))
+ return -1;
+
+ rte_strlcpy(name, params, RTE_CRYPTODEV_NAME_LEN);
+ dev = rte_cryptodev_pmd_get_named_dev(name);
+ if (!dev) {
+ plt_err("No cryptodev of name %s available", name);
+ return -1;
+ }
+
+ vf = dev->data->dev_private;
+
+ rte_tel_data_start_dict(d);
+ CNXK_TEL_DICT_INT(d, dev->data, dev_id);
+ CNXK_TEL_DICT_INT(d, dev->data, socket_id);
+ CNXK_TEL_DICT_INT(d, dev->data, dev_started);
+
+ /* Crypto capabilities */
+ crypto_caps = rte_tel_data_alloc();
+ crypto_caps_n = crypto_caps_array(crypto_caps, vf->crypto_caps,
+ CNXK_CPT_MAX_CAPS);
+ rte_tel_data_add_dict_container(d, "crypto_caps", crypto_caps, 0);
+ rte_tel_data_add_dict_int(d, "crypto_caps_
[dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
If the user requests to use an lcore above 128 using -l,
the eal will exit with "EAL: invalid core list syntax" and
very little else useful information.
THis patch adds some extra information suggesting to use --lcores
so that physical cores above RTE_MAX_LCORE (default 128) can be
used. This is achieved by using the --lcores option by mapping
the logical cores in the application to physical cores.
There is no change in functionalty, just additional messages
suggesting how the --lcores option might be used for the supplied
list of lcores. For example, if "-l 12-16,130,132" is used, we
see the following additional output on the command line:
EAL: Error = One of the 7 cores provided exceeds RTE_MAX_LCORE (128)
EAL: Please use --lcores instead, e.g.
--lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132
Signed-off-by: David Hunt
---
changes in v2
* Rather than increasing the default max lcores (as in v1),
it was agreed to do this instead (switch to --lcores).
* As the other patches in the v1 of the set are no longer related
to this change, I'll submit as a separate patch set.
changes in v3
* separated out some of the corelist cheking into separate function
* added extra messages for the different failure conditions.
* changed allocation of the core strings from static to dynamic
* now prints out a message for each core above RTE_MAX_LCORE
---
lib/eal/common/eal_common_options.c | 102
1 file changed, 89 insertions(+), 13 deletions(-)
diff --git a/lib/eal/common/eal_common_options.c
b/lib/eal/common/eal_common_options.c
index ff5861b5f3..6139b2b21e 100644
--- a/lib/eal/common/eal_common_options.c
+++ b/lib/eal/common/eal_common_options.c
@@ -709,6 +709,47 @@ update_lcore_config(int *cores)
return ret;
}
+static int
+check_core_list(int *lcores, unsigned int count)
+{
+ unsigned int i, j;
+ char *lcorestr;
+ int len = 0;
+ bool overflow = false;
+
+ for (j = 0; j < count; j++) {
+ if (lcores[j] >= RTE_MAX_LCORE) {
+ RTE_LOG(ERR, EAL, "lcore %d >= RTE_MAX_LCORE (%d)\n",
+ lcores[j], RTE_MAX_LCORE);
+ overflow = true;
+ }
+ }
+ if (overflow) {
+ /*
+* If we've encountered a core that's greater than
+* RTE_MAX_LCORE, suggest using --lcores option to
+* map lcores onto physical cores greater than
+* RTE_MAX_LCORE, then return.
+*/
+ lcorestr = calloc(1, RTE_MAX_LCORE * 10);
+ if (lcorestr == NULL) {
+ RTE_LOG(ERR, EAL, "Unable to allocate lcore string\n");
+ return -ENOMEM;
+ }
+ for (i = 0; i < count; i++)
+ len = len + snprintf(&lcorestr[len],
+ RTE_MAX_LCORE * 10 - len,
+ "%d@%d,", i, lcores[i]);
+ if (len > 0)
+ lcorestr[len-1] = 0;
+ RTE_LOG(ERR, EAL, "Please use --lcores instead, e.g. --lcores
%s\n",
+ lcorestr);
+ free(lcorestr);
+ return -1;
+ }
+ return 0;
+}
+
static int
eal_parse_coremask(const char *coremask, int *cores)
{
@@ -839,54 +880,89 @@ eal_parse_service_corelist(const char *corelist)
static int
eal_parse_corelist(const char *corelist, int *cores)
{
- unsigned count = 0;
+ unsigned int count = 0, k;
char *end = NULL;
int min, max;
int idx;
+ int lcores[RTE_MAX_LCORE];
+ char *corelist_copy;
for (idx = 0; idx < RTE_MAX_LCORE; idx++)
cores[idx] = -1;
+ corelist_copy = calloc(1, strlen(corelist)+1);
+ if (corelist_copy == NULL) {
+ RTE_LOG(ERR, EAL, "Unable to allocate corelist copy\n");
+ return -ENOMEM;
+ }
+
+ strlcpy(corelist_copy, corelist, strlen(corelist)+1);
+
/* Remove all blank characters ahead */
while (isblank(*corelist))
corelist++;
/* Get list of cores */
- min = RTE_MAX_LCORE;
+ min = -1;
do {
while (isblank(*corelist))
corelist++;
if (*corelist == '\0')
- return -1;
+ goto err;
errno = 0;
idx = strtol(corelist, &end, 10);
if (errno || end == NULL)
- return -1;
- if (idx < 0 || idx >= RTE_MAX_LCORE)
- return -1;
+ goto err;
+ if (idx < 0)
+ goto err;
while (isblank(*end))
end++;
if (*end == '-') {
min = idx;
} el
[dpdk-dev] [PATCH v3 2/2] eal: add additional info if core mask too long
If the user requests to use an lcore above 128 using -c,
the eal will exit with "EAL: invalid core list syntax" and
very little else useful information.
This patch adds some extra information suggesting to use --lcores
so that physical cores above RTE_MAX_LCORE (default 128) can be
used. This is achieved by using the --lcores option by mapping
the logical cores in the application to physical cores.
There is no change in functionalty, just additional messages
suggesting how the --lcores option might be used for the supplied
list of lcores. For example, if
"-c 0xf" is used, we see the
following additional output on the command line:
EAL: Error = One of the 4 cores provided exceeds RTE_MAX_LCORE (128)
EAL: Please use --lcores instead, e.g. --lcores 0@128,1@129,2@130,3@131
Signed-off-by: David Hunt
---
changes in v3
* added this patch to the set. Addresses the changes for
the -c option.
---
lib/eal/common/eal_common_options.c | 67 +++--
1 file changed, 55 insertions(+), 12 deletions(-)
diff --git a/lib/eal/common/eal_common_options.c
b/lib/eal/common/eal_common_options.c
index 6139b2b21e..331dca1e38 100644
--- a/lib/eal/common/eal_common_options.c
+++ b/lib/eal/common/eal_common_options.c
@@ -753,15 +753,25 @@ check_core_list(int *lcores, unsigned int count)
static int
eal_parse_coremask(const char *coremask, int *cores)
{
- unsigned count = 0;
+ unsigned int count = 0, k;
int i, j, idx;
int val;
char c;
+ int lcores[RTE_MAX_LCORE];
+ char *coremask_copy = NULL;
for (idx = 0; idx < RTE_MAX_LCORE; idx++)
cores[idx] = -1;
idx = 0;
+ coremask_copy = calloc(1, strlen(coremask)+1);
+ if (coremask_copy == NULL) {
+ RTE_LOG(ERR, EAL, "Unable to allocate coremask copy\n");
+ return -ENOMEM;
+ }
+
+ strlcpy(coremask_copy, coremask, strlen(coremask)+1);
+
/* Remove all blank characters ahead and after .
* Remove 0x/0X if exists.
*/
@@ -773,30 +783,63 @@ eal_parse_coremask(const char *coremask, int *cores)
i = strlen(coremask);
while ((i > 0) && isblank(coremask[i - 1]))
i--;
- if (i == 0)
- return -1;
+ if (i == 0) {
+ RTE_LOG(ERR, EAL, "No lcores in coremask: %s\n", coremask_copy);
+ goto err;
+ }
- for (i = i - 1; i >= 0 && idx < RTE_MAX_LCORE; i--) {
+ for (i = i - 1; i >= 0; i--) {
c = coremask[i];
if (isxdigit(c) == 0) {
/* invalid characters */
- return -1;
+ RTE_LOG(ERR, EAL, "invalid characters in coremask:
%s\n",
+ coremask_copy);
+ goto err;
}
val = xdigit2val(c);
- for (j = 0; j < BITS_PER_HEX && idx < RTE_MAX_LCORE; j++, idx++)
+ for (j = 0; j < BITS_PER_HEX; j++, idx++)
{
if ((1 << j) & val) {
- cores[idx] = count;
- count++;
+ lcores[count++] = idx;
+ if (count > RTE_MAX_LCORE) {
+ RTE_LOG(ERR, EAL, "Too many lcores
provided. Cannot exceed %d\n",
+ RTE_MAX_LCORE);
+ goto err;
+ }
}
}
}
for (; i >= 0; i--)
- if (coremask[i] != '0')
- return -1;
- if (count == 0)
- return -1;
+ if (coremask[i] != '0') {
+ RTE_LOG(ERR, EAL, "invalid characters in coremask:
%s\n",
+ coremask_copy);
+ goto err;
+ }
+ if (count == 0) {
+ RTE_LOG(ERR, EAL, "No lcores in coremask: %s\n", coremask_copy);
+ goto err;
+ }
+
+ if (check_core_list(lcores, count))
+ goto err;
+
+ /*
+* Now that we've gto a list of cores no longer than
+* RTE_MAX_LCORE, and no lcore in that list is greater
+* than RTE_MAX_LCORE, populate the cores
+* array and return.
+*/
+ for (k = 0; k < count; k++)
+ cores[lcores[k]] = k;
+
+ if (coremask_copy)
+ free(coremask_copy);
+
return 0;
+err:
+ if (coremask_copy)
+ free(coremask_copy);
+ return -1;
}
static int
--
2.17.1
Re: [dpdk-dev] [v8, 4/4] net/cnxk: add telemetry endpoing to ethdev
On Tue, Sep 21, 2021 at 04:57:52PM +0530, Jerin Jacob wrote:
> On Tue, Sep 21, 2021 at 4:23 PM Gowrishankar Muthukrishnan
> wrote:
> >
> > Add telemetry endpoint to ethdev.
> >
> > Signed-off-by: Gowrishankar Muthukrishnan
> > ---
> > drivers/net/cnxk/cnxk_ethdev_telemetry.c | 129 +++
> > drivers/net/cnxk/meson.build | 1 +
> > 2 files changed, 130 insertions(+)
> > create mode 100644 drivers/net/cnxk/cnxk_ethdev_telemetry.c
> >
> > diff --git a/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > new file mode 100644
> > index 00..de8c468670
> > --- /dev/null
> > +++ b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > @@ -0,0 +1,129 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(C) 2021 Marvell International Ltd.
> > + */
> > +
> > +#include
> > +
> > +#include "cnxk_ethdev.h"
> > +
> > +/* Macro to count no of words in eth_info_s size */
> > +#define ETH_INFO_SZ
> > \
> > + (RTE_ALIGN_CEIL(sizeof(struct eth_info_s), sizeof(uint64_t)) /
> >\
> > +sizeof(uint64_t))
> > +#define MACADDR_LEN 18
> > +
> > +static int
> > +ethdev_tel_handle_info(const char *cmd __rte_unused,
> > + const char *params __rte_unused, struct rte_tel_data
> > *d)
> > +{
> > + struct rte_eth_dev *eth_dev;
> > + struct rte_tel_data *i_data;
> > + struct cnxk_eth_dev *dev;
> > + union eth_info_u {
> > + struct eth_info_s {
> > + /** PF/VF information */
> > + uint16_t pf_func;
> > + /** No of rx queues */
> > + uint16_t rx_queues;
> > + /** No of tx queues */
> > + uint16_t tx_queues;
> > + /** Port ID */
> > + uint16_t port_id;
> > + /** MAC entries */
> > + char mac_addr[MACADDR_LEN];
> > + uint8_t max_mac_entries;
> > + bool dmac_filter_ena;
> > + uint8_t dmac_filter_count;
> > + uint16_t flags;
> > + uint8_t ptype_disable;
> > + bool scalar_ena;
> > + bool ptp_ena;
> > + /* Offload capabilities */
> > + uint64_t rx_offload_capa;
> > + uint64_t tx_offload_capa;
> > + uint32_t speed_capa;
> > + /* Configured offloads */
> > + uint64_t rx_offloads;
> > + uint64_t tx_offloads;
> > + /* Platform specific offload flags */
> > + uint16_t rx_offload_flags;
> > + uint16_t tx_offload_flags;
> > + /* ETHDEV RSS HF bitmask */
> > + uint64_t ethdev_rss_hf;
>
> + Ethdev, , mempool and telemetry maintainers
>
> All of the above data except[1] is generic data that can be derived
> from ethdev APIs or ethdev data from lib/ethdev itself,
> IMO, We should avoid duplicating this in driver and make useful for
> other driver/application by adding generic endpoint in ethdev.
> Same comment for "[2/4] mempool/cnxk: add telemetry end points" for
> mempool subsystem.
>
> Thoughts from Maintainers?
>
Not a maintainer of those libs, but +1 to making anything generic that can
be, save reimplementing things multiple times in drivers.
Re: [dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
On Tue, Sep 21, 2021 at 12:50:14PM +0100, David Hunt wrote: > If the user requests to use an lcore above 128 using -l, > the eal will exit with "EAL: invalid core list syntax" and > very little else useful information. > > THis patch adds some extra information suggesting to use --lcores > so that physical cores above RTE_MAX_LCORE (default 128) can be > used. This is achieved by using the --lcores option by mapping > the logical cores in the application to physical cores. > > There is no change in functionalty, just additional messages > suggesting how the --lcores option might be used for the supplied > list of lcores. For example, if "-l 12-16,130,132" is used, we > see the following additional output on the command line: > > EAL: Error = One of the 7 cores provided exceeds RTE_MAX_LCORE (128) > EAL: Please use --lcores instead, e.g. Minor suggestion: it would be good to clarify how to use lcores and what is happening here in the example. Something like: "Please use --lcores instead, to map lower lcore ids onto higher-numbered cores", could help the user understand better what is happening. > --lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132 > > Signed-off-by: David Hunt With some more info to help the user: Acked-by: Bruce Richardson
Re: [dpdk-dev] [PATCH v3 2/2] eal: add additional info if core mask too long
On Tue, Sep 21, 2021 at 12:50:15PM +0100, David Hunt wrote: > If the user requests to use an lcore above 128 using -c, > the eal will exit with "EAL: invalid core list syntax" and > very little else useful information. > > This patch adds some extra information suggesting to use --lcores > so that physical cores above RTE_MAX_LCORE (default 128) can be > used. This is achieved by using the --lcores option by mapping > the logical cores in the application to physical cores. > > There is no change in functionalty, just additional messages > suggesting how the --lcores option might be used for the supplied > list of lcores. For example, if > "-c 0xf" is used, we see the > following additional output on the command line: > > EAL: Error = One of the 4 cores provided exceeds RTE_MAX_LCORE (128) "=" ? > EAL: Please use --lcores instead, e.g. --lcores 0@128,1@129,2@130,3@131 > > Signed-off-by: David Hunt > While I really can't see many people using "-c" to use cores >128 (unless from a script, I suppose), this change is worthwhile having for completeness. I suggest aligning the error message with that for "-l" flag. Acked-by: Bruce Richardson
Re: [dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
On 21/9/2021 12:57 PM, Bruce Richardson wrote: On Tue, Sep 21, 2021 at 12:50:14PM +0100, David Hunt wrote: If the user requests to use an lcore above 128 using -l, the eal will exit with "EAL: invalid core list syntax" and very little else useful information. THis patch adds some extra information suggesting to use --lcores so that physical cores above RTE_MAX_LCORE (default 128) can be used. This is achieved by using the --lcores option by mapping the logical cores in the application to physical cores. There is no change in functionalty, just additional messages suggesting how the --lcores option might be used for the supplied list of lcores. For example, if "-l 12-16,130,132" is used, we see the following additional output on the command line: EAL: Error = One of the 7 cores provided exceeds RTE_MAX_LCORE (128) EAL: Please use --lcores instead, e.g. Minor suggestion: it would be good to clarify how to use lcores and what is happening here in the example. Something like: "Please use --lcores instead, to map lower lcore ids onto higher-numbered cores", could help the user understand better what is happening. Hi Bruce, how about: EAL: Please use --lcores to map logical cores onto cores > RTE_LCORE_MAX ,e.g. --lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132 Rgds, Dave. --lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132 Signed-off-by: David Hunt With some more info to help the user: Acked-by: Bruce Richardson
Re: [dpdk-dev] [v8, 4/4] net/cnxk: add telemetry endpoing to ethdev
On Tue, Sep 21, 2021 at 12:53:53PM +0100, Bruce Richardson wrote:
> On Tue, Sep 21, 2021 at 04:57:52PM +0530, Jerin Jacob wrote:
> > On Tue, Sep 21, 2021 at 4:23 PM Gowrishankar Muthukrishnan
> > wrote:
> > >
> > > Add telemetry endpoint to ethdev.
> > >
> > > Signed-off-by: Gowrishankar Muthukrishnan
> > > ---
> > > drivers/net/cnxk/cnxk_ethdev_telemetry.c | 129 +++
> > > drivers/net/cnxk/meson.build | 1 +
> > > 2 files changed, 130 insertions(+)
> > > create mode 100644 drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > >
> > > diff --git a/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > > b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > > new file mode 100644
> > > index 00..de8c468670
> > > --- /dev/null
> > > +++ b/drivers/net/cnxk/cnxk_ethdev_telemetry.c
> > > @@ -0,0 +1,129 @@
> > > +/* SPDX-License-Identifier: BSD-3-Clause
> > > + * Copyright(C) 2021 Marvell International Ltd.
> > > + */
> > > +
> > > +#include
> > > +
> > > +#include "cnxk_ethdev.h"
> > > +
> > > +/* Macro to count no of words in eth_info_s size */
> > > +#define ETH_INFO_SZ
> > > \
> > > + (RTE_ALIGN_CEIL(sizeof(struct eth_info_s), sizeof(uint64_t)) /
> > > \
> > > +sizeof(uint64_t))
> > > +#define MACADDR_LEN 18
> > > +
> > > +static int
> > > +ethdev_tel_handle_info(const char *cmd __rte_unused,
> > > + const char *params __rte_unused, struct
> > > rte_tel_data *d)
> > > +{
> > > + struct rte_eth_dev *eth_dev;
> > > + struct rte_tel_data *i_data;
> > > + struct cnxk_eth_dev *dev;
> > > + union eth_info_u {
> > > + struct eth_info_s {
> > > + /** PF/VF information */
> > > + uint16_t pf_func;
> > > + /** No of rx queues */
> > > + uint16_t rx_queues;
> > > + /** No of tx queues */
> > > + uint16_t tx_queues;
> > > + /** Port ID */
> > > + uint16_t port_id;
> > > + /** MAC entries */
> > > + char mac_addr[MACADDR_LEN];
> > > + uint8_t max_mac_entries;
> > > + bool dmac_filter_ena;
> > > + uint8_t dmac_filter_count;
> > > + uint16_t flags;
> > > + uint8_t ptype_disable;
> > > + bool scalar_ena;
> > > + bool ptp_ena;
> > > + /* Offload capabilities */
> > > + uint64_t rx_offload_capa;
> > > + uint64_t tx_offload_capa;
> > > + uint32_t speed_capa;
> > > + /* Configured offloads */
> > > + uint64_t rx_offloads;
> > > + uint64_t tx_offloads;
> > > + /* Platform specific offload flags */
> > > + uint16_t rx_offload_flags;
> > > + uint16_t tx_offload_flags;
> > > + /* ETHDEV RSS HF bitmask */
> > > + uint64_t ethdev_rss_hf;
> >
> > + Ethdev, , mempool and telemetry maintainers
> >
> > All of the above data except[1] is generic data that can be derived
> > from ethdev APIs or ethdev data from lib/ethdev itself,
> > IMO, We should avoid duplicating this in driver and make useful for
> > other driver/application by adding generic endpoint in ethdev.
> > Same comment for "[2/4] mempool/cnxk: add telemetry end points" for
> > mempool subsystem.
> >
> > Thoughts from Maintainers?
> >
> Not a maintainer of those libs, but +1 to making anything generic that can
> be, save reimplementing things multiple times in drivers.
I agree, I feel it would make more sense to have most of the job done in the
mempool library in generic. A new ops could be added to fill driver-specific
info.
Re: [dpdk-dev] [PATCH v2 02/20] net/ice/base: init imem table for parser
> -Original Message-
> From: Zhang, Qi Z
> Sent: Friday, September 17, 2021 10:43 PM
> To: Yang, Qiming
> Cc: Guo, Junfeng ; dev@dpdk.org; Zhang, Qi Z
>
> Subject: [PATCH v2 02/20] net/ice/base: init imem table for parser
>
> Parse DDP section ICE_SID_RXPARSER_IMEM into an array of struct
> ice_imem_item.
>
> Signed-off-by: Qi Zhang
> ---
.
> +/**
> + * ice_parser_sect_item_get - parse a item from a section
> + * @sect_type: section type
> + * @section: section object
> + * @index: index of the item to get
> + * @offset: dummy as prototype of ice_pkg_enum_entry's last parameter
> +*/ void *ice_parser_sect_item_get(u32 sect_type, void *section,
> +u32 index, u32 *offset)
> +{
> + struct ice_pkg_sect_hdr *hdr;
> + int data_off = ICE_SEC_DATA_OFFSET;
> + int size;
> +
> + if (!section)
> + return NULL;
> +
> + switch (sect_type) {
> + case ICE_SID_RXPARSER_IMEM:
> + size = ICE_SID_RXPARSER_IMEM_ENTRY_SIZE;
> + break;
> + default:
> + return NULL;
> + }
> +
> + hdr = (struct ice_pkg_sect_hdr *)section;
> + if (index >= LE16_TO_CPU(hdr->count))
> + return NULL;
> +
> + return (void *)((u64)section + data_off + index * size); }
Compile warning for 32bit, replace u64 with uintptr_t
> +
> +/**
> + * ice_parser_create_table - create a item table from a section
> + * @hw: pointer to the hardware structure
> + * @sect_type: section type
> + * @item_size: item size in byte
> + * @length: number of items in the table to create
> + * @item_get: the function will be parsed to ice_pkg_enum_entry
> + * @parser_item: the function to parse the item */ void
> +*ice_parser_create_table(struct ice_hw *hw, u32 sect_type,
> + u32 item_size, u32 length,
> + void *(*item_get)(u32 sect_type, void *section,
> + u32 index, u32 *offset),
> + void (*parse_item)(struct ice_hw *hw, u16 idx,
> + void *item, void *data,
> + int size))
> +{
> + struct ice_seg *seg = hw->seg;
> + struct ice_pkg_enum state;
> + u16 idx = 0;
> + void *table;
> + void *data;
> +
> + if (!seg)
> + return NULL;
> +
> + table = ice_malloc(hw, item_size * length);
> + if (!table) {
> + ice_debug(hw, ICE_DBG_PARSER, "failed to allocate memory for
> table type %d.\n",
> + sect_type);
> + return NULL;
> + }
> +
> + ice_memset(&state, 0, sizeof(state), ICE_NONDMA_MEM);
> + do {
> + data = ice_pkg_enum_entry(seg, &state, sect_type, NULL,
> + item_get);
> + seg = NULL;
> + if (data) {
> + struct ice_pkg_sect_hdr *hdr =
> + (struct ice_pkg_sect_hdr *)state.sect;
> +
> + idx = hdr->offset + state.entry_idx;
> + parse_item(hw, idx,
> +(void *)((u64)table + idx * item_size),
Same issue as above.
Will fix during apply.
[dpdk-dev] [PATCH v3 00/20] ice/base: add parser module
Add the parser module that can parse on a raw packet then figure out the low-level metadata to program the hardware packet process pipeline for flow offloading(Switch/FDIR/RSS). This is the pre-step to enable a protocol-agnostic flow offloading solution for ice devices that leverage Intel DDP technology. -v3: 1. fix 32 bit compile issue in patch 2/20 -v2: 1. use inclusive word in patch 2/20 2. replace magic number with macro in patch 17/20 3. fix couple typos Qi Zhang (20): net/ice/base: add parser create and destroy skeleton net/ice/base: init imem table for parser net/ice/base: init metainit table for parser net/ice/base: init parse graph cam table for parser net/ice/base: init boost TCAM table for parser net/ice/base: init ptype marker TCAM table for parser net/ice/base: init marker group table for parser net/ice/base: init protocol group table for parser net/ice/base: init flag redirect table for parser net/ice/base: init XLT key builder for parser net/ice/base: add parser runtime skeleton net/ice/base: add helper function for boost TCAM match net/ice/base: add helper functions for parse graph key matching net/ice/base: add helper function for ptype markers match net/ice/base: add helper function to redirect flags net/ice/base: add helper function to aggregate flags net/ice/base: add parser execution main loop net/ice/base: support double VLAN mode configure for parser net/ice/base: add tunnel port support for parser net/ice/base: add API for parser profile initialization drivers/net/ice/base/ice_bst_tcam.c| 291 + drivers/net/ice/base/ice_bst_tcam.h| 35 + drivers/net/ice/base/ice_common.h | 1 + drivers/net/ice/base/ice_flex_pipe.c | 4 +- drivers/net/ice/base/ice_flex_pipe.h | 8 + drivers/net/ice/base/ice_flex_type.h | 2 + drivers/net/ice/base/ice_flg_rd.c | 76 +++ drivers/net/ice/base/ice_flg_rd.h | 17 + drivers/net/ice/base/ice_imem.c| 244 +++ drivers/net/ice/base/ice_imem.h| 109 drivers/net/ice/base/ice_metainit.c| 143 drivers/net/ice/base/ice_metainit.h| 46 ++ drivers/net/ice/base/ice_mk_grp.c | 55 ++ drivers/net/ice/base/ice_mk_grp.h | 15 + drivers/net/ice/base/ice_parser.c | 556 drivers/net/ice/base/ice_parser.h | 113 drivers/net/ice/base/ice_parser_rt.c | 870 + drivers/net/ice/base/ice_parser_rt.h | 53 ++ drivers/net/ice/base/ice_parser_util.h | 36 + drivers/net/ice/base/ice_pg_cam.c | 374 +++ drivers/net/ice/base/ice_pg_cam.h | 74 +++ drivers/net/ice/base/ice_proto_grp.c | 108 +++ drivers/net/ice/base/ice_proto_grp.h | 23 + drivers/net/ice/base/ice_ptype_mk.c| 76 +++ drivers/net/ice/base/ice_ptype_mk.h| 21 + drivers/net/ice/base/ice_tmatch.h | 44 ++ drivers/net/ice/base/ice_type.h| 1 + drivers/net/ice/base/ice_xlt_kb.c | 216 ++ drivers/net/ice/base/ice_xlt_kb.h | 34 + drivers/net/ice/base/meson.build | 11 + 30 files changed, 3654 insertions(+), 2 deletions(-) create mode 100644 drivers/net/ice/base/ice_bst_tcam.c create mode 100644 drivers/net/ice/base/ice_bst_tcam.h create mode 100644 drivers/net/ice/base/ice_flg_rd.c create mode 100644 drivers/net/ice/base/ice_flg_rd.h create mode 100644 drivers/net/ice/base/ice_imem.c create mode 100644 drivers/net/ice/base/ice_imem.h create mode 100644 drivers/net/ice/base/ice_metainit.c create mode 100644 drivers/net/ice/base/ice_metainit.h create mode 100644 drivers/net/ice/base/ice_mk_grp.c create mode 100644 drivers/net/ice/base/ice_mk_grp.h create mode 100644 drivers/net/ice/base/ice_parser.c create mode 100644 drivers/net/ice/base/ice_parser.h create mode 100644 drivers/net/ice/base/ice_parser_rt.c create mode 100644 drivers/net/ice/base/ice_parser_rt.h create mode 100644 drivers/net/ice/base/ice_parser_util.h create mode 100644 drivers/net/ice/base/ice_pg_cam.c create mode 100644 drivers/net/ice/base/ice_pg_cam.h create mode 100644 drivers/net/ice/base/ice_proto_grp.c create mode 100644 drivers/net/ice/base/ice_proto_grp.h create mode 100644 drivers/net/ice/base/ice_ptype_mk.c create mode 100644 drivers/net/ice/base/ice_ptype_mk.h create mode 100644 drivers/net/ice/base/ice_tmatch.h create mode 100644 drivers/net/ice/base/ice_xlt_kb.c create mode 100644 drivers/net/ice/base/ice_xlt_kb.h -- 2.26.2
[dpdk-dev] [PATCH v3 01/20] net/ice/base: add parser create and destroy skeleton
Add new parser module which can parse a packet in binary
and generate information like ptype, protocol/offset pairs
and flags which can be used to feed the FXP profile creation
directly.
The patch added skeleton of the parser instance create and
destroy APIs:
ice_parser_create
ice_parser_destroy
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_common.h| 1 +
drivers/net/ice/base/ice_flex_pipe.c | 2 +-
drivers/net/ice/base/ice_flex_pipe.h | 5
drivers/net/ice/base/ice_parser.c| 34
drivers/net/ice/base/ice_parser.h| 14
drivers/net/ice/base/meson.build | 1 +
6 files changed, 56 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ice/base/ice_parser.c
create mode 100644 drivers/net/ice/base/ice_parser.h
diff --git a/drivers/net/ice/base/ice_common.h
b/drivers/net/ice/base/ice_common.h
index 1d8882c279..a3cbf4fb05 100644
--- a/drivers/net/ice/base/ice_common.h
+++ b/drivers/net/ice/base/ice_common.h
@@ -8,6 +8,7 @@
#include "ice_type.h"
#include "ice_nvm.h"
#include "ice_flex_pipe.h"
+#include "ice_parser.h"
#include "ice_switch.h"
#include "ice_fdir.h"
diff --git a/drivers/net/ice/base/ice_flex_pipe.c
b/drivers/net/ice/base/ice_flex_pipe.c
index 3631ddba2c..703c3e0416 100644
--- a/drivers/net/ice/base/ice_flex_pipe.c
+++ b/drivers/net/ice/base/ice_flex_pipe.c
@@ -284,7 +284,7 @@ ice_pkg_enum_section(struct ice_seg *ice_seg, struct
ice_pkg_enum *state,
* indicates a base offset of 10, and the index for the entry is 2, then
* section handler function should set the offset to 10 + 2 = 12.
*/
-static void *
+void *
ice_pkg_enum_entry(struct ice_seg *ice_seg, struct ice_pkg_enum *state,
u32 sect_type, u32 *offset,
void *(*handler)(u32 sect_type, void *section,
diff --git a/drivers/net/ice/base/ice_flex_pipe.h
b/drivers/net/ice/base/ice_flex_pipe.h
index b690be75fc..045a77c607 100644
--- a/drivers/net/ice/base/ice_flex_pipe.h
+++ b/drivers/net/ice/base/ice_flex_pipe.h
@@ -94,4 +94,9 @@ void ice_pkg_buf_free(struct ice_hw *hw, struct ice_buf_build
*bld);
enum ice_status
ice_set_key(u8 *key, u16 size, u8 *val, u8 *upd, u8 *dc, u8 *nm, u16 off,
u16 len);
+void *
+ice_pkg_enum_entry(struct ice_seg *ice_seg, struct ice_pkg_enum *state,
+ u32 sect_type, u32 *offset,
+ void *(*handler)(u32 sect_type, void *section,
+ u32 index, u32 *offset));
#endif /* _ICE_FLEX_PIPE_H_ */
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
new file mode 100644
index 00..c08decaf0d
--- /dev/null
+++ b/drivers/net/ice/base/ice_parser.c
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+
+/**
+ * ice_parser_create - create a parser instance
+ * @hw: pointer to the hardware structure
+ * @psr: output parameter for a new parser instance be created
+ */
+enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr)
+{
+ struct ice_parser *p;
+
+ p = (struct ice_parser *)ice_malloc(hw, sizeof(struct ice_parser));
+
+ if (!p)
+ return ICE_ERR_NO_MEMORY;
+
+ p->hw = hw;
+
+ *psr = p;
+ return ICE_SUCCESS;
+}
+
+/**
+ * ice_parser_destroy - destroy a parser instance
+ * @psr: pointer to a parser instance
+ */
+void ice_parser_destroy(struct ice_parser *psr)
+{
+ ice_free(psr->hw, psr);
+}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
new file mode 100644
index 00..5964bf4e49
--- /dev/null
+++ b/drivers/net/ice/base/ice_parser.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#ifndef _ICE_PARSER_H_
+#define _ICE_PARSER_H_
+
+struct ice_parser {
+ struct ice_hw *hw; /* pointer to the hardware structure */
+};
+
+enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
+void ice_parser_destroy(struct ice_parser *psr);
+#endif /* _ICE_PARSER_H_ */
diff --git a/drivers/net/ice/base/meson.build b/drivers/net/ice/base/meson.build
index be9713dfa1..2b0af54a5c 100644
--- a/drivers/net/ice/base/meson.build
+++ b/drivers/net/ice/base/meson.build
@@ -15,6 +15,7 @@ sources = [
'ice_acl_ctrl.c',
'ice_vlan_mode.c',
'ice_ptp_hw.c',
+ 'ice_parser.c',
]
error_cflags = [
--
2.26.2
[dpdk-dev] [PATCH v3 02/20] net/ice/base: init imem table for parser
Parse DDP section ICE_SID_RXPARSER_IMEM into an array of
struct ice_imem_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_imem.c| 244 +
drivers/net/ice/base/ice_imem.h| 109 +++
drivers/net/ice/base/ice_parser.c | 100 ++
drivers/net/ice/base/ice_parser.h | 3 +
drivers/net/ice/base/ice_parser_util.h | 25 +++
drivers/net/ice/base/ice_type.h| 1 +
drivers/net/ice/base/meson.build | 1 +
7 files changed, 483 insertions(+)
create mode 100644 drivers/net/ice/base/ice_imem.c
create mode 100644 drivers/net/ice/base/ice_imem.h
create mode 100644 drivers/net/ice/base/ice_parser_util.h
diff --git a/drivers/net/ice/base/ice_imem.c b/drivers/net/ice/base/ice_imem.c
new file mode 100644
index 00..aefc7132eb
--- /dev/null
+++ b/drivers/net/ice/base/ice_imem.c
@@ -0,0 +1,244 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_IMEM_TABLE_SIZE 192
+
+static void _imem_bst_bm_dump(struct ice_hw *hw, struct ice_bst_main *bm)
+{
+ ice_info(hw, "boost main:\n");
+ ice_info(hw, "\tal0 = %d\n", bm->al0);
+ ice_info(hw, "\tal1 = %d\n", bm->al1);
+ ice_info(hw, "\tal2 = %d\n", bm->al2);
+ ice_info(hw, "\tpg = %d\n", bm->pg);
+}
+
+static void _imem_bst_kb_dump(struct ice_hw *hw, struct ice_bst_keybuilder *kb)
+{
+ ice_info(hw, "boost key builder:\n");
+ ice_info(hw, "\tpriority = %d\n", kb->priority);
+ ice_info(hw, "\ttsr_ctrl = %d\n", kb->tsr_ctrl);
+}
+
+static void _imem_np_kb_dump(struct ice_hw *hw, struct ice_np_keybuilder *kb)
+{
+ ice_info(hw, "next proto key builder:\n");
+ ice_info(hw, "\tops = %d\n", kb->ops);
+ ice_info(hw, "\tstart_or_reg0 = %d\n", kb->start_or_reg0);
+ ice_info(hw, "\tlen_or_reg1 = %d\n", kb->len_or_reg1);
+}
+
+static void _imem_pg_kb_dump(struct ice_hw *hw, struct ice_pg_keybuilder *kb)
+{
+ ice_info(hw, "parse graph key builder:\n");
+ ice_info(hw, "\tflag0_ena = %d\n", kb->flag0_ena);
+ ice_info(hw, "\tflag1_ena = %d\n", kb->flag1_ena);
+ ice_info(hw, "\tflag2_ena = %d\n", kb->flag2_ena);
+ ice_info(hw, "\tflag3_ena = %d\n", kb->flag3_ena);
+ ice_info(hw, "\tflag0_idx = %d\n", kb->flag0_idx);
+ ice_info(hw, "\tflag1_idx = %d\n", kb->flag1_idx);
+ ice_info(hw, "\tflag2_idx = %d\n", kb->flag2_idx);
+ ice_info(hw, "\tflag3_idx = %d\n", kb->flag3_idx);
+ ice_info(hw, "\talu_reg_idx = %d\n", kb->alu_reg_idx);
+}
+
+static void _imem_alu_dump(struct ice_hw *hw, struct ice_alu *alu, int index)
+{
+ ice_info(hw, "alu%d:\n", index);
+ ice_info(hw, "\topc = %d\n", alu->opc);
+ ice_info(hw, "\tsrc_start = %d\n", alu->src_start);
+ ice_info(hw, "\tsrc_len = %d\n", alu->src_len);
+ ice_info(hw, "\tshift_xlate_select = %d\n", alu->shift_xlate_select);
+ ice_info(hw, "\tshift_xlate_key = %d\n", alu->shift_xlate_key);
+ ice_info(hw, "\tsrc_reg_id = %d\n", alu->src_reg_id);
+ ice_info(hw, "\tdst_reg_id = %d\n", alu->dst_reg_id);
+ ice_info(hw, "\tinc0 = %d\n", alu->inc0);
+ ice_info(hw, "\tinc1 = %d\n", alu->inc1);
+ ice_info(hw, "\tproto_offset_opc = %d\n", alu->proto_offset_opc);
+ ice_info(hw, "\tproto_offset = %d\n", alu->proto_offset);
+ ice_info(hw, "\tbranch_addr = %d\n", alu->branch_addr);
+ ice_info(hw, "\timm = %d\n", alu->imm);
+ ice_info(hw, "\tdst_start = %d\n", alu->dst_start);
+ ice_info(hw, "\tdst_len = %d\n", alu->dst_len);
+ ice_info(hw, "\tflags_extr_imm = %d\n", alu->flags_extr_imm);
+ ice_info(hw, "\tflags_start_imm= %d\n", alu->flags_start_imm);
+}
+
+/**
+ * ice_imem_dump - dump an imem item info
+ * @ice_hw: pointer to the hardware structure
+ * @item: imem item to dump
+ */
+void ice_imem_dump(struct ice_hw *hw, struct ice_imem_item *item)
+{
+ ice_info(hw, "index = %d\n", item->idx);
+ _imem_bst_bm_dump(hw, &item->b_m);
+ _imem_bst_kb_dump(hw, &item->b_kb);
+ ice_info(hw, "pg priority = %d\n", item->pg);
+ _imem_np_kb_dump(hw, &item->np_kb);
+ _imem_pg_kb_dump(hw, &item->pg_kb);
+ _imem_alu_dump(hw, &item->alu0, 0);
+ _imem_alu_dump(hw, &item->alu1, 1);
+ _imem_alu_dump(hw, &item->alu2, 2);
+}
+
+/** The function parses a 4 bits Boost Main with below format:
+ * BIT 0: ALU 0 (bm->alu0)
+ * BIT 1: ALU 1 (bm->alu1)
+ * BIT 2: ALU 2 (bm->alu2)
+ * BIT 3: Parge Graph (bm->pg)
+ */
+static void _imem_bm_init(struct ice_bst_main *bm, u8 data)
+{
+ bm->al0 = (data & 0x1) != 0;
+ bm->al1 = (data & 0x2) != 0;
+ bm->al2 = (data & 0x4) != 0;
+ bm->pg = (data & 0x8) != 0;
+}
+
+/** The function parses a 10 bits Boost Main Build with below format:
+ * BIT 0-7: Priority (bkb->priority)
+ * BIT 8: TSR Control (bkb->tsr_ctrl)
+ *
[dpdk-dev] [PATCH v3 03/20] net/ice/base: init metainit table for parser
Parse DDP section ICE_SID_RXPARSER_METADATA_INIT into an array of
struct ice_metainit_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_metainit.c| 143 +
drivers/net/ice/base/ice_metainit.h| 46
drivers/net/ice/base/ice_parser.c | 15 ++-
drivers/net/ice/base/ice_parser.h | 2 +
drivers/net/ice/base/ice_parser_util.h | 1 +
drivers/net/ice/base/meson.build | 1 +
6 files changed, 206 insertions(+), 2 deletions(-)
create mode 100644 drivers/net/ice/base/ice_metainit.c
create mode 100644 drivers/net/ice/base/ice_metainit.h
diff --git a/drivers/net/ice/base/ice_metainit.c
b/drivers/net/ice/base/ice_metainit.c
new file mode 100644
index 00..5d49c6861d
--- /dev/null
+++ b/drivers/net/ice/base/ice_metainit.c
@@ -0,0 +1,143 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_METAINIT_TABLE_SIZE 16
+
+/**
+ * ice_metainit_dump - dump an metainit item info
+ * @ice_hw: pointer to the hardware structure
+ * @item: metainit item to dump
+ */
+void ice_metainit_dump(struct ice_hw *hw, struct ice_metainit_item *item)
+{
+ ice_info(hw, "index = %d\n", item->idx);
+ ice_info(hw, "tsr = %d\n", item->tsr);
+ ice_info(hw, "ho = %d\n", item->ho);
+ ice_info(hw, "pc = %d\n", item->pc);
+ ice_info(hw, "pg_rn = %d\n", item->pg_rn);
+ ice_info(hw, "cd = %d\n", item->cd);
+ ice_info(hw, "gpr_a_ctrl = %d\n", item->gpr_a_ctrl);
+ ice_info(hw, "gpr_a_data_mdid = %d\n", item->gpr_a_data_mdid);
+ ice_info(hw, "gpr_a_data_start = %d\n", item->gpr_a_data_start);
+ ice_info(hw, "gpr_a_data_len = %d\n", item->gpr_a_data_len);
+ ice_info(hw, "gpr_a_id = %d\n", item->gpr_a_id);
+ ice_info(hw, "gpr_b_ctrl = %d\n", item->gpr_b_ctrl);
+ ice_info(hw, "gpr_b_data_mdid = %d\n", item->gpr_b_data_mdid);
+ ice_info(hw, "gpr_b_data_start = %d\n", item->gpr_b_data_start);
+ ice_info(hw, "gpr_b_data_len = %d\n", item->gpr_b_data_len);
+ ice_info(hw, "gpr_b_id = %d\n", item->gpr_b_id);
+ ice_info(hw, "gpr_c_ctrl = %d\n", item->gpr_c_ctrl);
+ ice_info(hw, "gpr_c_data_mdid = %d\n", item->gpr_c_data_mdid);
+ ice_info(hw, "gpr_c_data_start = %d\n", item->gpr_c_data_start);
+ ice_info(hw, "gpr_c_data_len = %d\n", item->gpr_c_data_len);
+ ice_info(hw, "gpr_c_id = %d\n", item->gpr_c_id);
+ ice_info(hw, "gpr_d_ctrl = %d\n", item->gpr_d_ctrl);
+ ice_info(hw, "gpr_d_data_mdid = %d\n", item->gpr_d_data_mdid);
+ ice_info(hw, "gpr_d_data_start = %d\n", item->gpr_d_data_start);
+ ice_info(hw, "gpr_d_data_len = %d\n", item->gpr_d_data_len);
+ ice_info(hw, "gpr_d_id = %d\n", item->gpr_d_id);
+ ice_info(hw, "flags = 0x%016" PRIx64 "\n", item->flags);
+}
+
+/** The function parses a 192 bits Metadata Init entry with below format:
+ * BIT 0-7: TCAM Search Key Register (mi->tsr)
+ * BIT 8-16: Header Offset (mi->ho)
+ * BIT 17-24: Program Counter (mi->pc)
+ * BIT 25-35: Parse Graph Root Node (mi->pg_rn)
+ * BIT 36-38: Control Domain (mi->cd)
+ * BIT 39:GPR_A Data Control (mi->gpr_a_ctrl)
+ * BIT 40-44: GPR_A MDID.ID (mi->gpr_a_data_mdid)
+ * BIT 45-48: GPR_A MDID.START (mi->gpr_a_data_start)
+ * BIT 49-53: GPR_A MDID.LEN (mi->gpr_a_data_len)
+ * BIT 54-55: reserved
+ * BIT 56-59: GPR_A ID (mi->gpr_a_id)
+ * BIT 60:GPR_B Data Control (mi->gpr_b_ctrl)
+ * BIT 61-65: GPR_B MDID.ID (mi->gpr_b_data_mdid)
+ * BIT 66-69: GPR_B MDID.START (mi->gpr_b_data_start)
+ * BIT 70-74: GPR_B MDID.LEN (mi->gpr_b_data_len)
+ * BIT 75-76: reserved
+ * BIT 77-80: GPR_B ID (mi->gpr_a_id)
+ * BIT 81:GPR_C Data Control (mi->gpr_c_ctrl)
+ * BIT 82-86: GPR_C MDID.ID (mi->gpr_c_data_mdid)
+ * BIT 87-90: GPR_C MDID.START (mi->gpr_c_data_start)
+ * BIT 91-95: GPR_C MDID.LEN (mi->gpr_c_data_len)
+ * BIT 96-97: reserved
+ * BIT 98-101:GPR_C ID (mi->gpr_c_id)
+ * BIT 102: GPR_D Data Control (mi->gpr_d_ctrl)
+ * BIT 103-107:GPR_D MDID.ID (mi->gpr_d_data_mdid)
+ * BIT 108-111:GPR_D MDID.START (mi->gpr_d_data_start)
+ * BIT 112-116:GPR_D MDID.LEN (mi->gpr_d_data_len)
+ * BIT 117-118:reserved
+ * BIT 119-122:GPR_D ID (mi->gpr_d_id)
+ * BIT 123-186:Flags (mi->flags)
+ * BIT 187-191:rserved
+ */
+static void _metainit_parse_item(struct ice_hw *hw, u16 idx, void *item,
+void *data, int size)
+{
+ struct ice_metainit_item *mi = (struct ice_metainit_item *)item;
+ u8 *buf = (u8 *)data;
+ u64 d64;
+
+ mi->idx = idx;
+ d64 = *(u64 *)buf;
+
+ mi->tsr = (u8)(d64 & 0xff);
+ mi->ho = (u16)((d64 >> 8) & 0x1ff);
+ mi->pc = (u16)((d64 >> 17) & 0xff);
+ mi->pg_rn = (u16)((d64 >> 25) & 0x3ff);
+ mi->cd = (u16)((d64 >> 36) & 0x7);
+ mi->gpr_a_ctrl = ((d64 >> 39) & 0x1) != 0;
+ mi->gpr_a_
Re: [dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
On 21/9/2021 1:04 PM, David Hunt wrote: On 21/9/2021 12:57 PM, Bruce Richardson wrote: On Tue, Sep 21, 2021 at 12:50:14PM +0100, David Hunt wrote: If the user requests to use an lcore above 128 using -l, the eal will exit with "EAL: invalid core list syntax" and very little else useful information. THis patch adds some extra information suggesting to use --lcores so that physical cores above RTE_MAX_LCORE (default 128) can be used. This is achieved by using the --lcores option by mapping the logical cores in the application to physical cores. There is no change in functionalty, just additional messages suggesting how the --lcores option might be used for the supplied list of lcores. For example, if "-l 12-16,130,132" is used, we see the following additional output on the command line: EAL: Error = One of the 7 cores provided exceeds RTE_MAX_LCORE (128) EAL: Please use --lcores instead, e.g. Minor suggestion: it would be good to clarify how to use lcores and what is happening here in the example. Something like: "Please use --lcores instead, to map lower lcore ids onto higher-numbered cores", could help the user understand better what is happening. Hi Bruce, how about: EAL: Please use --lcores to map logical cores onto cores > RTE_LCORE_MAX ,e.g. --lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132 Rgds, Dave. I think this should do it, as it clarifies the mapping: EAL: lcore 130 >= RTE_MAX_LCORE (128) EAL: lcore 132 >= RTE_MAX_LCORE (128) EAL: to use high physical core ids , please use --lcores to map them to lcore ids below RTE_LCORE_MAX, e.g.'--lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132' Thanks, Dave. --lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132 Signed-off-by: David Hunt With some more info to help the user: Acked-by: Bruce Richardson
[dpdk-dev] [PATCH v3 04/20] net/ice/base: init parse graph cam table for parser
Parse DDP section ICE_SID_RXPARSER_CAM or ICE_SID_RXPARSER_PG_SPILL
into an array of struct ice_pg_cam_item.
Parse DDP section ICE_SID_RXPARSER_NOMATCH_CAM or
ICE_SID_RXPARSER_NOMATCH_SPILL into an array of struct ice_pg_nm_cam_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser.c | 44 +
drivers/net/ice/base/ice_parser.h | 12 ++
drivers/net/ice/base/ice_pg_cam.c | 298 ++
drivers/net/ice/base/ice_pg_cam.h | 68 +++
drivers/net/ice/base/meson.build | 1 +
5 files changed, 423 insertions(+)
create mode 100644 drivers/net/ice/base/ice_pg_cam.c
create mode 100644 drivers/net/ice/base/ice_pg_cam.h
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 0dd6bf137d..f6a1c821f1 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -8,6 +8,10 @@
#define ICE_SEC_DATA_OFFSET4
#define ICE_SID_RXPARSER_IMEM_ENTRY_SIZE 48
#define ICE_SID_RXPARSER_METADATA_INIT_ENTRY_SIZE 24
+#define ICE_SID_RXPARSER_CAM_ENTRY_SIZE16
+#define ICE_SID_RXPARSER_PG_SPILL_ENTRY_SIZE 17
+#define ICE_SID_RXPARSER_NOMATCH_CAM_ENTRY_SIZE12
+#define ICE_SID_RXPARSER_NOMATCH_SPILL_ENTRY_SIZE 13
/**
* ice_parser_sect_item_get - parse a item from a section
@@ -33,6 +37,18 @@ void *ice_parser_sect_item_get(u32 sect_type, void *section,
case ICE_SID_RXPARSER_METADATA_INIT:
size = ICE_SID_RXPARSER_METADATA_INIT_ENTRY_SIZE;
break;
+ case ICE_SID_RXPARSER_CAM:
+ size = ICE_SID_RXPARSER_CAM_ENTRY_SIZE;
+ break;
+ case ICE_SID_RXPARSER_PG_SPILL:
+ size = ICE_SID_RXPARSER_PG_SPILL_ENTRY_SIZE;
+ break;
+ case ICE_SID_RXPARSER_NOMATCH_CAM:
+ size = ICE_SID_RXPARSER_NOMATCH_CAM_ENTRY_SIZE;
+ break;
+ case ICE_SID_RXPARSER_NOMATCH_SPILL:
+ size = ICE_SID_RXPARSER_NOMATCH_SPILL_ENTRY_SIZE;
+ break;
default:
return NULL;
}
@@ -125,6 +141,30 @@ enum ice_status ice_parser_create(struct ice_hw *hw,
struct ice_parser **psr)
goto err;
}
+ p->pg_cam_table = ice_pg_cam_table_get(hw);
+ if (!p->pg_cam_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
+ p->pg_sp_cam_table = ice_pg_sp_cam_table_get(hw);
+ if (!p->pg_sp_cam_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
+ p->pg_nm_cam_table = ice_pg_nm_cam_table_get(hw);
+ if (!p->pg_nm_cam_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
+ p->pg_nm_sp_cam_table = ice_pg_nm_sp_cam_table_get(hw);
+ if (!p->pg_nm_sp_cam_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
*psr = p;
return ICE_SUCCESS;
err:
@@ -140,6 +180,10 @@ void ice_parser_destroy(struct ice_parser *psr)
{
ice_free(psr->hw, psr->imem_table);
ice_free(psr->hw, psr->mi_table);
+ ice_free(psr->hw, psr->pg_cam_table);
+ ice_free(psr->hw, psr->pg_sp_cam_table);
+ ice_free(psr->hw, psr->pg_nm_cam_table);
+ ice_free(psr->hw, psr->pg_nm_sp_cam_table);
ice_free(psr->hw, psr);
}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index b7d0b23ded..b157e27510 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -5,6 +5,10 @@
#ifndef _ICE_PARSER_H_
#define _ICE_PARSER_H_
+#include "ice_metainit.h"
+#include "ice_imem.h"
+#include "ice_pg_cam.h"
+
struct ice_parser {
struct ice_hw *hw; /* pointer to the hardware structure */
@@ -12,6 +16,14 @@ struct ice_parser {
struct ice_imem_item *imem_table;
/* load data from section ICE_SID_RXPARSER_METADATA_INIT */
struct ice_metainit_item *mi_table;
+ /* load data from section ICE_SID_RXPARSER_CAM */
+ struct ice_pg_cam_item *pg_cam_table;
+ /* load data from section ICE_SID_RXPARSER_PG_SPILL */
+ struct ice_pg_cam_item *pg_sp_cam_table;
+ /* load data from section ICE_SID_RXPARSER_NOMATCH_CAM */
+ struct ice_pg_nm_cam_item *pg_nm_cam_table;
+ /* load data from section ICE_SID_RXPARSER_NOMATCH_SPILL */
+ struct ice_pg_nm_cam_item *pg_nm_sp_cam_table;
};
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
diff --git a/drivers/net/ice/base/ice_pg_cam.c
b/drivers/net/ice/base/ice_pg_cam.c
new file mode 100644
index 00..171986bf3d
--- /dev/null
+++ b/drivers/net/ice/base/ice_pg_cam.c
@@ -0,0 +1,298 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+static void _pg_cam_key_dump(struct
[dpdk-dev] [PATCH v3 05/20] net/ice/base: init boost TCAM table for parser
Parse DDP section ICE_SID_RXPARSER_CAM into an array of
ice_bst_tcam_item.
Parse DDP section ICE_SID_LBL_RXPARSER_TMEM into an array of
ice_lbl_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_bst_tcam.c| 241 +
drivers/net/ice/base/ice_bst_tcam.h| 28 +++
drivers/net/ice/base/ice_imem.c| 2 +-
drivers/net/ice/base/ice_metainit.c| 2 +-
drivers/net/ice/base/ice_parser.c | 48 -
drivers/net/ice/base/ice_parser.h | 5 +
drivers/net/ice/base/ice_parser_util.h | 12 +-
drivers/net/ice/base/ice_pg_cam.c | 8 +-
drivers/net/ice/base/meson.build | 1 +
9 files changed, 337 insertions(+), 10 deletions(-)
create mode 100644 drivers/net/ice/base/ice_bst_tcam.c
create mode 100644 drivers/net/ice/base/ice_bst_tcam.h
diff --git a/drivers/net/ice/base/ice_bst_tcam.c
b/drivers/net/ice/base/ice_bst_tcam.c
new file mode 100644
index 00..1c82359681
--- /dev/null
+++ b/drivers/net/ice/base/ice_bst_tcam.c
@@ -0,0 +1,241 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_BST_TCAM_TABLE_SIZE 256
+
+static void _bst_np_kb_dump(struct ice_hw *hw, struct ice_np_keybuilder *kb)
+{
+ ice_info(hw, "next proto key builder:\n");
+ ice_info(hw, "\tops = %d\n", kb->ops);
+ ice_info(hw, "\tstart_or_reg0 = %d\n", kb->start_or_reg0);
+ ice_info(hw, "\tlen_or_reg1 = %d\n", kb->len_or_reg1);
+}
+
+static void _bst_pg_kb_dump(struct ice_hw *hw, struct ice_pg_keybuilder *kb)
+{
+ ice_info(hw, "parse graph key builder:\n");
+ ice_info(hw, "\tflag0_ena = %d\n", kb->flag0_ena);
+ ice_info(hw, "\tflag1_ena = %d\n", kb->flag1_ena);
+ ice_info(hw, "\tflag2_ena = %d\n", kb->flag2_ena);
+ ice_info(hw, "\tflag3_ena = %d\n", kb->flag3_ena);
+ ice_info(hw, "\tflag0_idx = %d\n", kb->flag0_idx);
+ ice_info(hw, "\tflag1_idx = %d\n", kb->flag1_idx);
+ ice_info(hw, "\tflag2_idx = %d\n", kb->flag2_idx);
+ ice_info(hw, "\tflag3_idx = %d\n", kb->flag3_idx);
+ ice_info(hw, "\talu_reg_idx = %d\n", kb->alu_reg_idx);
+}
+
+static void _bst_alu_dump(struct ice_hw *hw, struct ice_alu *alu, int index)
+{
+ ice_info(hw, "alu%d:\n", index);
+ ice_info(hw, "\topc = %d\n", alu->opc);
+ ice_info(hw, "\tsrc_start = %d\n", alu->src_start);
+ ice_info(hw, "\tsrc_len = %d\n", alu->src_len);
+ ice_info(hw, "\tshift_xlate_select = %d\n", alu->shift_xlate_select);
+ ice_info(hw, "\tshift_xlate_key = %d\n", alu->shift_xlate_key);
+ ice_info(hw, "\tsrc_reg_id = %d\n", alu->src_reg_id);
+ ice_info(hw, "\tdst_reg_id = %d\n", alu->dst_reg_id);
+ ice_info(hw, "\tinc0 = %d\n", alu->inc0);
+ ice_info(hw, "\tinc1 = %d\n", alu->inc1);
+ ice_info(hw, "\tproto_offset_opc = %d\n", alu->proto_offset_opc);
+ ice_info(hw, "\tproto_offset = %d\n", alu->proto_offset);
+ ice_info(hw, "\tbranch_addr = %d\n", alu->branch_addr);
+ ice_info(hw, "\timm = %d\n", alu->imm);
+ ice_info(hw, "\tdst_start = %d\n", alu->dst_start);
+ ice_info(hw, "\tdst_len = %d\n", alu->dst_len);
+ ice_info(hw, "\tflags_extr_imm = %d\n", alu->flags_extr_imm);
+ ice_info(hw, "\tflags_start_imm= %d\n", alu->flags_start_imm);
+}
+
+/**
+ * ice_bst_tcam_dump - dump a boost tcam info
+ * @ice_hw: pointer to the hardware structure
+ * @item: boost tcam to dump
+ */
+void ice_bst_tcam_dump(struct ice_hw *hw, struct ice_bst_tcam_item *item)
+{
+ int i;
+
+ ice_info(hw, "address = %d\n", item->address);
+ ice_info(hw, "key:");
+ for (i = 0; i < 20; i++)
+ ice_info(hw, "%02x ", item->key[i]);
+ ice_info(hw, "\n");
+ ice_info(hw, "key_inv:");
+ for (i = 0; i < 20; i++)
+ ice_info(hw, "%02x ", item->key_inv[i]);
+ ice_info(hw, "\n");
+ ice_info(hw, "hit_idx_grp = %d\n", item->hit_idx_grp);
+ ice_info(hw, "pg_pri = %d\n", item->pg_pri);
+ _bst_np_kb_dump(hw, &item->np_kb);
+ _bst_pg_kb_dump(hw, &item->pg_kb);
+ _bst_alu_dump(hw, &item->alu0, 0);
+ _bst_alu_dump(hw, &item->alu1, 1);
+ _bst_alu_dump(hw, &item->alu2, 2);
+}
+
+/** The function parses a 96 bits ALU entry with below format:
+ * BIT 0-5: Opcode (alu->opc)
+ * BIT 6-13: Source Start (alu->src_start)
+ * BIT 14-18: Source Length (alu->src_len)
+ * BIT 19:Shift/Xlate Select (alu->shift_xlate_select)
+ * BIT 20-23: Shift/Xlate Key (alu->shift_xlate_key)
+ * BIT 24-30: Source Register ID (alu->src_reg_id)
+ * BIT 31-37: Dest. Register ID (alu->dst_reg_id)
+ * BIT 38:Inc0 (alu->inc0)
+ * BIT 39:Inc1:(alu->inc1)
+ * BIT 40:41 Protocol Offset Opcode (alu->proto_offset_opc)
+ * BIT 42:49 Protocol Offset (alu->proto_offset)
+ * BIT 50:57 Branch Address (alu->branch_addr)
+ * BIT 58:73 Immediate (alu->imm)
+ *
[dpdk-dev] [PATCH v3 06/20] net/ice/base: init ptype marker TCAM table for parser
Parse DDP section ICE_SID_RXPARSER_MARKER_PTYPE into an array of
ice_ptype_mk_tcam_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser.c | 11 ++
drivers/net/ice/base/ice_parser.h | 3 ++
drivers/net/ice/base/ice_ptype_mk.c | 54 +
drivers/net/ice/base/ice_ptype_mk.h | 18 ++
drivers/net/ice/base/meson.build| 1 +
5 files changed, 87 insertions(+)
create mode 100644 drivers/net/ice/base/ice_ptype_mk.c
create mode 100644 drivers/net/ice/base/ice_ptype_mk.h
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 594f802800..e7b79453a8 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -13,6 +13,7 @@
#define ICE_SID_RXPARSER_NOMATCH_CAM_ENTRY_SIZE12
#define ICE_SID_RXPARSER_NOMATCH_SPILL_ENTRY_SIZE 13
#define ICE_SID_RXPARSER_BOOST_TCAM_ENTRY_SIZE 88
+#define ICE_SID_RXPARSER_MARKER_TYPE_ENTRY_SIZE24
#define ICE_SEC_LBL_DATA_OFFSET2
#define ICE_SID_LBL_ENTRY_SIZE 66
@@ -72,6 +73,9 @@ void *ice_parser_sect_item_get(u32 sect_type, void *section,
data_off = ICE_SEC_LBL_DATA_OFFSET;
size = ICE_SID_LBL_ENTRY_SIZE;
break;
+ case ICE_SID_RXPARSER_MARKER_PTYPE:
+ size = ICE_SID_RXPARSER_MARKER_TYPE_ENTRY_SIZE;
+ break;
default:
return NULL;
}
@@ -205,6 +209,12 @@ enum ice_status ice_parser_create(struct ice_hw *hw,
struct ice_parser **psr)
goto err;
}
+ p->ptype_mk_tcam_table = ice_ptype_mk_tcam_table_get(hw);
+ if (!p->ptype_mk_tcam_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
*psr = p;
return ICE_SUCCESS;
err:
@@ -226,6 +236,7 @@ void ice_parser_destroy(struct ice_parser *psr)
ice_free(psr->hw, psr->pg_nm_sp_cam_table);
ice_free(psr->hw, psr->bst_tcam_table);
ice_free(psr->hw, psr->bst_lbl_table);
+ ice_free(psr->hw, psr->ptype_mk_tcam_table);
ice_free(psr->hw, psr);
}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index 319648970a..40b7b823c3 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -9,6 +9,7 @@
#include "ice_imem.h"
#include "ice_pg_cam.h"
#include "ice_bst_tcam.h"
+#include "ice_ptype_mk.h"
struct ice_parser {
struct ice_hw *hw; /* pointer to the hardware structure */
@@ -29,6 +30,8 @@ struct ice_parser {
struct ice_bst_tcam_item *bst_tcam_table;
/* load data from section ICE_SID_LBL_RXPARSER_TMEM */
struct ice_lbl_item *bst_lbl_table;
+ /* load data from section ICE_SID_RXPARSER_MARKER_PTYPE */
+ struct ice_ptype_mk_tcam_item *ptype_mk_tcam_table;
};
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
diff --git a/drivers/net/ice/base/ice_ptype_mk.c
b/drivers/net/ice/base/ice_ptype_mk.c
new file mode 100644
index 00..33623dcfbc
--- /dev/null
+++ b/drivers/net/ice/base/ice_ptype_mk.c
@@ -0,0 +1,54 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_PTYPE_MK_TCAM_TABLE_SIZE 1024
+
+/**
+ * ice_ptype_mk_tcam_dump - dump an ptype marker tcam info_
+ * @ice_hw: pointer to the hardware structure
+ * @item: ptype marker tcam to dump
+ */
+void ice_ptype_mk_tcam_dump(struct ice_hw *hw,
+ struct ice_ptype_mk_tcam_item *item)
+{
+ int i;
+
+ ice_info(hw, "address = %d\n", item->address);
+ ice_info(hw, "ptype = %d\n", item->ptype);
+ ice_info(hw, "key:");
+ for (i = 0; i < 10; i++)
+ ice_info(hw, "%02x ", item->key[i]);
+ ice_info(hw, "\n");
+ ice_info(hw, "key_inv:");
+ for (i = 0; i < 10; i++)
+ ice_info(hw, "%02x ", item->key_inv[i]);
+ ice_info(hw, "\n");
+}
+
+static void _parse_ptype_mk_tcam_item(struct ice_hw *hw, u16 idx, void *item,
+ void *data, int size)
+{
+ ice_parse_item_dflt(hw, idx, item, data, size);
+
+ if (hw->debug_mask & ICE_DBG_PARSER)
+ ice_ptype_mk_tcam_dump(hw,
+ (struct ice_ptype_mk_tcam_item *)item);
+}
+
+/**
+ * ice_ptype_mk_tcam_table_get - create a ptype marker tcam table
+ * @ice_hw: pointer to the hardware structure
+ */
+struct ice_ptype_mk_tcam_item *ice_ptype_mk_tcam_table_get(struct ice_hw *hw)
+{
+ return (struct ice_ptype_mk_tcam_item *)
+ ice_parser_create_table(hw, ICE_SID_RXPARSER_MARKER_PTYPE,
+ sizeof(struct ice_ptype_mk_tcam_item),
+ ICE_PTYPE_MK_TCAM_TABLE_SIZE,
+
[dpdk-dev] [PATCH v3 07/20] net/ice/base: init marker group table for parser
Parse DDP section ICE_SID_RXPARSER_MARKER_GRP into an array of
ice_mk_grp_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_mk_grp.c | 55 +++
drivers/net/ice/base/ice_mk_grp.h | 15 +
drivers/net/ice/base/ice_parser.c | 11 +++
drivers/net/ice/base/ice_parser.h | 3 ++
drivers/net/ice/base/meson.build | 1 +
5 files changed, 85 insertions(+)
create mode 100644 drivers/net/ice/base/ice_mk_grp.c
create mode 100644 drivers/net/ice/base/ice_mk_grp.h
diff --git a/drivers/net/ice/base/ice_mk_grp.c
b/drivers/net/ice/base/ice_mk_grp.c
new file mode 100644
index 00..4e9ab5c13a
--- /dev/null
+++ b/drivers/net/ice/base/ice_mk_grp.c
@@ -0,0 +1,55 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_MK_GRP_TABLE_SIZE 128
+#define ICE_MK_COUNT_PER_GRP 8
+
+/**
+ * ice_mk_grp_dump - dump an marker group item info
+ * @ice_hw: pointer to the hardware structure
+ * @item: marker group item to dump
+ */
+void ice_mk_grp_dump(struct ice_hw *hw, struct ice_mk_grp_item *item)
+{
+ int i;
+
+ ice_info(hw, "index = %d\n", item->idx);
+ ice_info(hw, "markers: ");
+ for (i = 0; i < ICE_MK_COUNT_PER_GRP; i++)
+ ice_info(hw, "%d ", item->markers[i]);
+ ice_info(hw, "\n");
+}
+
+static void _mk_grp_parse_item(struct ice_hw *hw, u16 idx, void *item,
+ void *data, int size)
+{
+ struct ice_mk_grp_item *grp = (struct ice_mk_grp_item *)item;
+ u8 *buf = (u8 *)data;
+ int i;
+
+ grp->idx = idx;
+
+ for (i = 0; i < ICE_MK_COUNT_PER_GRP; i++)
+ grp->markers[i] = buf[i];
+
+ if (hw->debug_mask & ICE_DBG_PARSER)
+ ice_mk_grp_dump(hw, grp);
+}
+
+/**
+ * ice_mk_grp_table_get - create a marker group table
+ * @ice_hw: pointer to the hardware structure
+ */
+struct ice_mk_grp_item *ice_mk_grp_table_get(struct ice_hw *hw)
+{
+ return (struct ice_mk_grp_item *)
+ ice_parser_create_table(hw, ICE_SID_RXPARSER_MARKER_GRP,
+ sizeof(struct ice_mk_grp_item),
+ ICE_MK_GRP_TABLE_SIZE,
+ ice_parser_sect_item_get,
+ _mk_grp_parse_item, false);
+}
diff --git a/drivers/net/ice/base/ice_mk_grp.h
b/drivers/net/ice/base/ice_mk_grp.h
new file mode 100644
index 00..04d11b49c2
--- /dev/null
+++ b/drivers/net/ice/base/ice_mk_grp.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#ifndef _ICE_MK_GRP_H_
+#define _ICE_MK_GRP_H_
+
+struct ice_mk_grp_item {
+ int idx;
+ u8 markers[8];
+};
+
+void ice_mk_grp_dump(struct ice_hw *hw, struct ice_mk_grp_item *item);
+struct ice_mk_grp_item *ice_mk_grp_table_get(struct ice_hw *hw);
+#endif /* _ICE_MK_GRP_H_ */
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index e7b79453a8..142da462ee 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -14,6 +14,7 @@
#define ICE_SID_RXPARSER_NOMATCH_SPILL_ENTRY_SIZE 13
#define ICE_SID_RXPARSER_BOOST_TCAM_ENTRY_SIZE 88
#define ICE_SID_RXPARSER_MARKER_TYPE_ENTRY_SIZE24
+#define ICE_SID_RXPARSER_MARKER_GRP_ENTRY_SIZE 8
#define ICE_SEC_LBL_DATA_OFFSET2
#define ICE_SID_LBL_ENTRY_SIZE 66
@@ -76,6 +77,9 @@ void *ice_parser_sect_item_get(u32 sect_type, void *section,
case ICE_SID_RXPARSER_MARKER_PTYPE:
size = ICE_SID_RXPARSER_MARKER_TYPE_ENTRY_SIZE;
break;
+ case ICE_SID_RXPARSER_MARKER_GRP:
+ size = ICE_SID_RXPARSER_MARKER_GRP_ENTRY_SIZE;
+ break;
default:
return NULL;
}
@@ -215,6 +219,12 @@ enum ice_status ice_parser_create(struct ice_hw *hw,
struct ice_parser **psr)
goto err;
}
+ p->mk_grp_table = ice_mk_grp_table_get(hw);
+ if (!p->mk_grp_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
*psr = p;
return ICE_SUCCESS;
err:
@@ -237,6 +247,7 @@ void ice_parser_destroy(struct ice_parser *psr)
ice_free(psr->hw, psr->bst_tcam_table);
ice_free(psr->hw, psr->bst_lbl_table);
ice_free(psr->hw, psr->ptype_mk_tcam_table);
+ ice_free(psr->hw, psr->mk_grp_table);
ice_free(psr->hw, psr);
}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index 40b7b823c3..b390eecb2b 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -10,6 +10,7 @@
#include "ice_pg_cam.h"
#include "ice_bst_tcam.h"
#include "ice_ptype_mk.h"
+#include "ice_mk_grp.h"
struct ice
[dpdk-dev] [PATCH v3 08/20] net/ice/base: init protocol group table for parser
Parse DDP section ICE_SID_RXPARSER_PROTO_GRP into an array of
ice_proto_grp_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser.c| 11 +++
drivers/net/ice/base/ice_parser.h| 3 +
drivers/net/ice/base/ice_proto_grp.c | 108 +++
drivers/net/ice/base/ice_proto_grp.h | 23 ++
drivers/net/ice/base/meson.build | 1 +
5 files changed, 146 insertions(+)
create mode 100644 drivers/net/ice/base/ice_proto_grp.c
create mode 100644 drivers/net/ice/base/ice_proto_grp.h
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 142da462ee..0ed7c5fe32 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -15,6 +15,7 @@
#define ICE_SID_RXPARSER_BOOST_TCAM_ENTRY_SIZE 88
#define ICE_SID_RXPARSER_MARKER_TYPE_ENTRY_SIZE24
#define ICE_SID_RXPARSER_MARKER_GRP_ENTRY_SIZE 8
+#define ICE_SID_RXPARSER_PROTO_GRP_ENTRY_SIZE 24
#define ICE_SEC_LBL_DATA_OFFSET2
#define ICE_SID_LBL_ENTRY_SIZE 66
@@ -80,6 +81,9 @@ void *ice_parser_sect_item_get(u32 sect_type, void *section,
case ICE_SID_RXPARSER_MARKER_GRP:
size = ICE_SID_RXPARSER_MARKER_GRP_ENTRY_SIZE;
break;
+ case ICE_SID_RXPARSER_PROTO_GRP:
+ size = ICE_SID_RXPARSER_PROTO_GRP_ENTRY_SIZE;
+ break;
default:
return NULL;
}
@@ -225,6 +229,12 @@ enum ice_status ice_parser_create(struct ice_hw *hw,
struct ice_parser **psr)
goto err;
}
+ p->proto_grp_table = ice_proto_grp_table_get(hw);
+ if (!p->proto_grp_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
*psr = p;
return ICE_SUCCESS;
err:
@@ -248,6 +258,7 @@ void ice_parser_destroy(struct ice_parser *psr)
ice_free(psr->hw, psr->bst_lbl_table);
ice_free(psr->hw, psr->ptype_mk_tcam_table);
ice_free(psr->hw, psr->mk_grp_table);
+ ice_free(psr->hw, psr->proto_grp_table);
ice_free(psr->hw, psr);
}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index b390eecb2b..ac4139abd7 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -11,6 +11,7 @@
#include "ice_bst_tcam.h"
#include "ice_ptype_mk.h"
#include "ice_mk_grp.h"
+#include "ice_proto_grp.h"
struct ice_parser {
struct ice_hw *hw; /* pointer to the hardware structure */
@@ -35,6 +36,8 @@ struct ice_parser {
struct ice_ptype_mk_tcam_item *ptype_mk_tcam_table;
/* load data from section ICE_SID_RXPARSER_MARKER_GRP */
struct ice_mk_grp_item *mk_grp_table;
+ /* load data from section ICE_SID_RXPARSER_PROTO_GRP */
+ struct ice_proto_grp_item *proto_grp_table;
};
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
diff --git a/drivers/net/ice/base/ice_proto_grp.c
b/drivers/net/ice/base/ice_proto_grp.c
new file mode 100644
index 00..69d5d9a18a
--- /dev/null
+++ b/drivers/net/ice/base/ice_proto_grp.c
@@ -0,0 +1,108 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_PROTO_GRP_TABLE_SIZE 192
+
+static void _proto_off_dump(struct ice_hw *hw, struct ice_proto_off *po,
+ int idx)
+{
+ ice_info(hw, "proto %d\n", idx);
+ ice_info(hw, "\tpolarity = %d\n", po->polarity);
+ ice_info(hw, "\tproto_id = %d\n", po->proto_id);
+ ice_info(hw, "\toffset = %d\n", po->offset);
+}
+
+/**
+ * ice_proto_grp_dump - dump a proto group item info
+ * @ice_hw: pointer to the hardware structure
+ * @item: proto group item to dump
+ */
+void ice_proto_grp_dump(struct ice_hw *hw, struct ice_proto_grp_item *item)
+{
+ int i;
+
+ ice_info(hw, "index = %d\n", item->idx);
+
+ for (i = 0; i < ICE_PROTO_COUNT_PER_GRP; i++)
+ _proto_off_dump(hw, &item->po[i], i);
+}
+
+/** The function parses a 22 bits Protocol entry with below format:
+ * BIT 0: Polarity of Protocol Offset (po->polarity)
+ * BIT 1-8: Protocol ID (po->proto_id)
+ * BIT 9-11: reserved
+ * BIT 12-21: Protocol Offset (po->offset)
+ */
+static void _proto_off_parse(struct ice_proto_off *po, u32 data)
+{
+ po->polarity = (data & 0x1) != 0;
+ po->proto_id = (u8)((data >> 1) & 0xff);
+ po->offset = (u16)((data >> 12) & 0x3ff);
+}
+
+/** The function parses a 192 bits Protocol Group Table entry with below
+ * format:
+ * BIT 0-21: Protocol 0 (grp->po[0])
+ * BIT 22-43: Protocol 1 (grp->po[1])
+ * BIT 44-65: Protocol 2 (grp->po[2])
+ * BIT 66-87: Protocol 3 (grp->po[3])
+ * BIT 88-109:Protocol 4 (grp->po[4])
+ * BIT 110-131:Protocol 5 (grp->po[5])
+ * BIT 132-153:Protocol 6 (grp->po[6])
+ * BI
[dpdk-dev] [PATCH v3 09/20] net/ice/base: init flag redirect table for parser
Parse DDP section ICE_SID_RXPARSER_FLAG_REDIR into an array of
ice_flag_rd_item.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_flex_type.h | 2 ++
drivers/net/ice/base/ice_flg_rd.c| 53
drivers/net/ice/base/ice_flg_rd.h| 16 +
drivers/net/ice/base/ice_parser.c| 11 ++
drivers/net/ice/base/ice_parser.h| 3 ++
drivers/net/ice/base/meson.build | 1 +
6 files changed, 86 insertions(+)
create mode 100644 drivers/net/ice/base/ice_flg_rd.c
create mode 100644 drivers/net/ice/base/ice_flg_rd.h
diff --git a/drivers/net/ice/base/ice_flex_type.h
b/drivers/net/ice/base/ice_flex_type.h
index 3f2038c931..59eeca0a30 100644
--- a/drivers/net/ice/base/ice_flex_type.h
+++ b/drivers/net/ice/base/ice_flex_type.h
@@ -198,6 +198,8 @@ struct ice_buf_hdr {
#define ICE_SID_CDID_KEY_BUILDER_PE87
#define ICE_SID_CDID_REDIR_PE 88
+#define ICE_SID_RXPARSER_FLAG_REDIR97
+
/* Label Metadata section IDs */
#define ICE_SID_LBL_FIRST 0x8010
#define ICE_SID_LBL_RXPARSER_IMEM 0x8010
diff --git a/drivers/net/ice/base/ice_flg_rd.c
b/drivers/net/ice/base/ice_flg_rd.c
new file mode 100644
index 00..292916d9a8
--- /dev/null
+++ b/drivers/net/ice/base/ice_flg_rd.c
@@ -0,0 +1,53 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+#include "ice_parser_util.h"
+
+#define ICE_FLG_RD_TABLE_SIZE 64
+
+/**
+ * ice_flg_rd_dump - dump a flag redirect item info
+ * @ice_hw: pointer to the hardware structure
+ * @item: flag redirect item to dump
+ */
+void ice_flg_rd_dump(struct ice_hw *hw, struct ice_flg_rd_item *item)
+{
+ ice_info(hw, "index = %d\n", item->idx);
+ ice_info(hw, "expose = %d\n", item->expose);
+ ice_info(hw, "intr_flg_id = %d\n", item->intr_flg_id);
+}
+
+/** The function parses a 8 bits Flag Redirect Table entry with below format:
+ * BIT 0: Expose (rdi->expose)
+ * BIT 1-6: Internal Flag ID (rdi->intr_flg_id)
+ * BIT 7: reserved
+ */
+static void _flg_rd_parse_item(struct ice_hw *hw, u16 idx, void *item,
+ void *data, int size)
+{
+ struct ice_flg_rd_item *rdi = (struct ice_flg_rd_item *)item;
+ u8 d8 = *(u8 *)data;
+
+ rdi->idx = idx;
+ rdi->expose = (d8 & 0x1) != 0;
+ rdi->intr_flg_id = (u8)((d8 >> 1) & 0x3f);
+
+ if (hw->debug_mask & ICE_DBG_PARSER)
+ ice_flg_rd_dump(hw, rdi);
+}
+
+/**
+ * ice_flg_rd_table_get - create a flag redirect table
+ * @ice_hw: pointer to the hardware structure
+ */
+struct ice_flg_rd_item *ice_flg_rd_table_get(struct ice_hw *hw)
+{
+ return (struct ice_flg_rd_item *)
+ ice_parser_create_table(hw, ICE_SID_RXPARSER_FLAG_REDIR,
+ sizeof(struct ice_flg_rd_item),
+ ICE_FLG_RD_TABLE_SIZE,
+ ice_parser_sect_item_get,
+ _flg_rd_parse_item, false);
+}
diff --git a/drivers/net/ice/base/ice_flg_rd.h
b/drivers/net/ice/base/ice_flg_rd.h
new file mode 100644
index 00..e65350f18c
--- /dev/null
+++ b/drivers/net/ice/base/ice_flg_rd.h
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#ifndef _ICE_FLG_RD_H_
+#define _ICE_FLG_RD_H_
+
+struct ice_flg_rd_item {
+ u16 idx;
+ bool expose;
+ u8 intr_flg_id;
+};
+
+void ice_flg_rd_dump(struct ice_hw *hw, struct ice_flg_rd_item *item);
+struct ice_flg_rd_item *ice_flg_rd_table_get(struct ice_hw *hw);
+#endif /* _ICE_FLG_RD_H_ */
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 0ed7c5fe32..02a299aa04 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -16,6 +16,7 @@
#define ICE_SID_RXPARSER_MARKER_TYPE_ENTRY_SIZE24
#define ICE_SID_RXPARSER_MARKER_GRP_ENTRY_SIZE 8
#define ICE_SID_RXPARSER_PROTO_GRP_ENTRY_SIZE 24
+#define ICE_SID_RXPARSER_FLAG_REDIR_ENTRY_SIZE 1
#define ICE_SEC_LBL_DATA_OFFSET2
#define ICE_SID_LBL_ENTRY_SIZE 66
@@ -84,6 +85,9 @@ void *ice_parser_sect_item_get(u32 sect_type, void *section,
case ICE_SID_RXPARSER_PROTO_GRP:
size = ICE_SID_RXPARSER_PROTO_GRP_ENTRY_SIZE;
break;
+ case ICE_SID_RXPARSER_FLAG_REDIR:
+ size = ICE_SID_RXPARSER_FLAG_REDIR_ENTRY_SIZE;
+ break;
default:
return NULL;
}
@@ -235,6 +239,12 @@ enum ice_status ice_parser_create(struct ice_hw *hw,
struct ice_parser **psr)
goto err;
}
+ p->flg_rd_table = ice_flg_rd_table_get(hw);
+ if (!p->flg_rd_table) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
*psr =
[dpdk-dev] [PATCH v3 10/20] net/ice/base: init XLT key builder for parser
Parse below DDP section into struct ice_xlt_kb:
ICE_SID_XLT_KEY_BUILDER_SW
ICE_SID_XLT_KEY_BUILDER_FD
ICE_SID_XLT_KEY_BUILDER_RSS
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_flex_pipe.c | 2 +-
drivers/net/ice/base/ice_flex_pipe.h | 3 +
drivers/net/ice/base/ice_parser.c| 28
drivers/net/ice/base/ice_parser.h| 9 ++
drivers/net/ice/base/ice_xlt_kb.c| 189 +++
drivers/net/ice/base/ice_xlt_kb.h| 33 +
drivers/net/ice/base/meson.build | 1 +
7 files changed, 264 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ice/base/ice_xlt_kb.c
create mode 100644 drivers/net/ice/base/ice_xlt_kb.h
diff --git a/drivers/net/ice/base/ice_flex_pipe.c
b/drivers/net/ice/base/ice_flex_pipe.c
index 703c3e0416..f35d59f4f5 100644
--- a/drivers/net/ice/base/ice_flex_pipe.c
+++ b/drivers/net/ice/base/ice_flex_pipe.c
@@ -218,7 +218,7 @@ ice_pkg_advance_sect(struct ice_seg *ice_seg, struct
ice_pkg_enum *state)
* When the function returns a NULL pointer, then the end of the matching
* sections has been reached.
*/
-static void *
+void *
ice_pkg_enum_section(struct ice_seg *ice_seg, struct ice_pkg_enum *state,
u32 sect_type)
{
diff --git a/drivers/net/ice/base/ice_flex_pipe.h
b/drivers/net/ice/base/ice_flex_pipe.h
index 045a77c607..9733c4b214 100644
--- a/drivers/net/ice/base/ice_flex_pipe.h
+++ b/drivers/net/ice/base/ice_flex_pipe.h
@@ -99,4 +99,7 @@ ice_pkg_enum_entry(struct ice_seg *ice_seg, struct
ice_pkg_enum *state,
u32 sect_type, u32 *offset,
void *(*handler)(u32 sect_type, void *section,
u32 index, u32 *offset));
+void *
+ice_pkg_enum_section(struct ice_seg *ice_seg, struct ice_pkg_enum *state,
+u32 sect_type);
#endif /* _ICE_FLEX_PIPE_H_ */
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 02a299aa04..1bce75a05c 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -245,6 +245,30 @@ enum ice_status ice_parser_create(struct ice_hw *hw,
struct ice_parser **psr)
goto err;
}
+ p->xlt_kb_sw = ice_xlt_kb_get_sw(hw);
+ if (!p->xlt_kb_sw) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
+ p->xlt_kb_acl = ice_xlt_kb_get_acl(hw);
+ if (!p->xlt_kb_acl) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
+ p->xlt_kb_fd = ice_xlt_kb_get_fd(hw);
+ if (!p->xlt_kb_fd) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
+ p->xlt_kb_rss = ice_xlt_kb_get_rss(hw);
+ if (!p->xlt_kb_rss) {
+ status = ICE_ERR_PARAM;
+ goto err;
+ }
+
*psr = p;
return ICE_SUCCESS;
err:
@@ -270,6 +294,10 @@ void ice_parser_destroy(struct ice_parser *psr)
ice_free(psr->hw, psr->mk_grp_table);
ice_free(psr->hw, psr->proto_grp_table);
ice_free(psr->hw, psr->flg_rd_table);
+ ice_free(psr->hw, psr->xlt_kb_sw);
+ ice_free(psr->hw, psr->xlt_kb_acl);
+ ice_free(psr->hw, psr->xlt_kb_fd);
+ ice_free(psr->hw, psr->xlt_kb_rss);
ice_free(psr->hw, psr);
}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index 272eb28d5b..ba3175e60f 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -13,6 +13,7 @@
#include "ice_mk_grp.h"
#include "ice_proto_grp.h"
#include "ice_flg_rd.h"
+#include "ice_xlt_kb.h"
struct ice_parser {
struct ice_hw *hw; /* pointer to the hardware structure */
@@ -41,6 +42,14 @@ struct ice_parser {
struct ice_proto_grp_item *proto_grp_table;
/* load data from section ICE_SID_RXPARSER_FLAG_REDIR */
struct ice_flg_rd_item *flg_rd_table;
+ /* load data from section ICE_SID_XLT_KEY_BUILDER_SW */
+ struct ice_xlt_kb *xlt_kb_sw;
+ /* load data from section ICE_SID_XLT_KEY_BUILDER_ACL */
+ struct ice_xlt_kb *xlt_kb_acl;
+ /* load data from section ICE_SID_XLT_KEY_BUILDER_FD */
+ struct ice_xlt_kb *xlt_kb_fd;
+ /* load data from section ICE_SID_XLT_KEY_BUILDER_RSS */
+ struct ice_xlt_kb *xlt_kb_rss;
};
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
diff --git a/drivers/net/ice/base/ice_xlt_kb.c
b/drivers/net/ice/base/ice_xlt_kb.c
new file mode 100644
index 00..8b4043a836
--- /dev/null
+++ b/drivers/net/ice/base/ice_xlt_kb.c
@@ -0,0 +1,189 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+
+#define ICE_XLT_KB_TBL_OFF 12
+#define ICE_XLT_KB_TBL_ENTRY_SIZE 24
+
+static void _xlt_kb_entry_dump(struct ice_hw *hw,
+ struct ice_xlt_kb_entry *entry, int idx)
+{
+ int i;
+
+ ice_info(hw, "key builder
[dpdk-dev] [PATCH v3 11/20] net/ice/base: add parser runtime skeleton
Add parser runtime data struct ice_parser_rt.
Add below APIs for parser runtime preparation:
ice_parser_rt_reset
ice_parser_rt_pkt_buf_set
Add below API skeleton for parser runtime execution:
ice_parser_rt_execute
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser.c| 39 +++
drivers/net/ice/base/ice_parser.h| 24 +++
drivers/net/ice/base/ice_parser_rt.c | 98
drivers/net/ice/base/ice_parser_rt.h | 28
drivers/net/ice/base/meson.build | 1 +
5 files changed, 190 insertions(+)
create mode 100644 drivers/net/ice/base/ice_parser_rt.c
create mode 100644 drivers/net/ice/base/ice_parser_rt.h
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 1bce75a05c..8f423710fc 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -167,6 +167,8 @@ enum ice_status ice_parser_create(struct ice_hw *hw, struct
ice_parser **psr)
struct ice_parser *p;
p = (struct ice_parser *)ice_malloc(hw, sizeof(struct ice_parser));
+ p->hw = hw;
+ p->rt.psr = p;
if (!p)
return ICE_ERR_NO_MEMORY;
@@ -301,3 +303,40 @@ void ice_parser_destroy(struct ice_parser *psr)
ice_free(psr->hw, psr);
}
+
+/**
+ * ice_parser_run - parse on a packet in binary and return the result
+ * @psr: pointer to a parser instance
+ * @pkt_buf: packet data
+ * @pkt_len: packet length
+ * @rslt: input/output parameter to save parser result.
+ */
+enum ice_status ice_parser_run(struct ice_parser *psr, const u8 *pkt_buf,
+ int pkt_len, struct ice_parser_result *rslt)
+{
+ ice_parser_rt_reset(&psr->rt);
+ ice_parser_rt_pktbuf_set(&psr->rt, pkt_buf, pkt_len);
+
+ return ice_parser_rt_execute(&psr->rt, rslt);
+}
+
+/**
+ * ice_parser_result_dump - dump a parser result info
+ * @hw: pointer to the hardware structure
+ * @rslt: parser result info to dump
+ */
+void ice_parser_result_dump(struct ice_hw *hw, struct ice_parser_result *rslt)
+{
+ int i;
+
+ ice_info(hw, "ptype = %d\n", rslt->ptype);
+ for (i = 0; i < rslt->po_num; i++)
+ ice_info(hw, "proto = %d, offset = %d\n",
+rslt->po[i].proto_id, rslt->po[i].offset);
+
+ ice_info(hw, "flags_psr = 0x%016" PRIx64 "\n", rslt->flags_psr);
+ ice_info(hw, "flags_pkt = 0x%016" PRIx64 "\n", rslt->flags_pkt);
+ ice_info(hw, "flags_sw = 0x%04x\n", rslt->flags_sw);
+ ice_info(hw, "flags_fd = 0x%04x\n", rslt->flags_fd);
+ ice_info(hw, "flags_rss = 0x%04x\n", rslt->flags_rss);
+}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index ba3175e60f..f71a8d8775 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -14,6 +14,7 @@
#include "ice_proto_grp.h"
#include "ice_flg_rd.h"
#include "ice_xlt_kb.h"
+#include "ice_parser_rt.h"
struct ice_parser {
struct ice_hw *hw; /* pointer to the hardware structure */
@@ -50,8 +51,31 @@ struct ice_parser {
struct ice_xlt_kb *xlt_kb_fd;
/* load data from section ICE_SID_XLT_KEY_BUILDER_RSS */
struct ice_xlt_kb *xlt_kb_rss;
+ struct ice_parser_rt rt; /* parser runtime */
};
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
void ice_parser_destroy(struct ice_parser *psr);
+
+struct ice_parser_proto_off {
+ u8 proto_id; /* hardware protocol ID */
+ u16 offset; /* offset where the protocol header start */
+};
+
+struct ice_parser_result {
+ u16 ptype; /* 16 bits hardware PTYPE */
+ /* protocol and header offset pairs */
+ struct ice_parser_proto_off po[16];
+ int po_num; /* number of pairs must <= 16 */
+ u64 flags_psr; /* 64 bits parser flags */
+ u64 flags_pkt; /* 64 bits packet flags */
+ u16 flags_sw; /* 16 bits key builder flag for SW */
+ u16 flags_acl; /* 16 bits key builder flag for ACL */
+ u16 flags_fd; /* 16 bits key builder flag for FD */
+ u16 flags_rss; /* 16 bits key builder flag for RSS */
+};
+
+enum ice_status ice_parser_run(struct ice_parser *psr, const u8 *pkt_buf,
+ int pkt_len, struct ice_parser_result *rslt);
+void ice_parser_result_dump(struct ice_hw *hw, struct ice_parser_result *rslt);
#endif /* _ICE_PARSER_H_ */
diff --git a/drivers/net/ice/base/ice_parser_rt.c
b/drivers/net/ice/base/ice_parser_rt.c
new file mode 100644
index 00..d62d0170e5
--- /dev/null
+++ b/drivers/net/ice/base/ice_parser_rt.c
@@ -0,0 +1,98 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#include "ice_common.h"
+
+#define GPR_HB_IDX 64
+#define GPR_ERR_IDX84
+#define GPR_FLG_IDX104
+#define GPR_TSR_IDX108
+#define GPR_NN_IDX 109
+#define GPR_HO_IDX 110
+#define GPR_NP_IDX 111
+
+static void _rt_tsr_set(struct ice
[dpdk-dev] [PATCH v3 12/20] net/ice/base: add helper function for boost TCAM match
Add internal helper function ice_bst_tcam_match to perform ternary
match on boost TCAM.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_bst_tcam.c | 22 +++
drivers/net/ice/base/ice_bst_tcam.h | 3 ++
drivers/net/ice/base/ice_parser.h | 1 +
drivers/net/ice/base/ice_tmatch.h | 44 +
4 files changed, 70 insertions(+)
create mode 100644 drivers/net/ice/base/ice_tmatch.h
diff --git a/drivers/net/ice/base/ice_bst_tcam.c
b/drivers/net/ice/base/ice_bst_tcam.c
index 1c82359681..76b3a5c551 100644
--- a/drivers/net/ice/base/ice_bst_tcam.c
+++ b/drivers/net/ice/base/ice_bst_tcam.c
@@ -239,3 +239,25 @@ struct ice_lbl_item *ice_bst_lbl_table_get(struct ice_hw
*hw)
ice_parser_sect_item_get,
_parse_lbl_item, true);
}
+
+/**
+ * ice_bst_tcam_match - match a pattern on the boost tcam table
+ * @tcam_table: boost tcam table to search
+ * @pat: pattern to match
+ */
+struct ice_bst_tcam_item *
+ice_bst_tcam_match(struct ice_bst_tcam_item *tcam_table, u8 *pat)
+{
+ int i;
+
+ for (i = 0; i < ICE_BST_TCAM_TABLE_SIZE; i++) {
+ struct ice_bst_tcam_item *item = &tcam_table[i];
+
+ if (item->hit_idx_grp == 0)
+ continue;
+ if (ice_ternary_match(item->key, item->key_inv, pat, 20))
+ return item;
+ }
+
+ return NULL;
+}
diff --git a/drivers/net/ice/base/ice_bst_tcam.h
b/drivers/net/ice/base/ice_bst_tcam.h
index a4ab40721f..3cba0bbf55 100644
--- a/drivers/net/ice/base/ice_bst_tcam.h
+++ b/drivers/net/ice/base/ice_bst_tcam.h
@@ -25,4 +25,7 @@ void ice_bst_tcam_dump(struct ice_hw *hw, struct
ice_bst_tcam_item *item);
struct ice_bst_tcam_item *ice_bst_tcam_table_get(struct ice_hw *hw);
struct ice_lbl_item *ice_bst_lbl_table_get(struct ice_hw *hw);
+
+struct ice_bst_tcam_item *
+ice_bst_tcam_match(struct ice_bst_tcam_item *tcam_table, u8 *pat);
#endif /*_ICE_BST_TCAM_H_ */
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index f71a8d8775..715158d6eb 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -15,6 +15,7 @@
#include "ice_flg_rd.h"
#include "ice_xlt_kb.h"
#include "ice_parser_rt.h"
+#include "ice_tmatch.h"
struct ice_parser {
struct ice_hw *hw; /* pointer to the hardware structure */
diff --git a/drivers/net/ice/base/ice_tmatch.h
b/drivers/net/ice/base/ice_tmatch.h
new file mode 100644
index 00..178a084639
--- /dev/null
+++ b/drivers/net/ice/base/ice_tmatch.h
@@ -0,0 +1,44 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2021 Intel Corporation
+ */
+
+#ifndef _ICE_TMATCH_H_
+#define _ICE_TMATCH_H_
+
+static inline
+bool ice_ternary_match_byte(u8 key, u8 key_inv, u8 pat)
+{
+ u8 k1, k2, v;
+ int i;
+
+ for (i = 0; i < 8; i++) {
+ k1 = (u8)(key & (1 << i));
+ k2 = (u8)(key_inv & (1 << i));
+ v = (u8)(pat & (1 << i));
+
+ if (k1 != 0 && k2 != 0)
+ continue;
+ if (k1 == 0 && k2 == 0)
+ return false;
+
+ if (k1 == v)
+ return false;
+ }
+
+ return true;
+}
+
+static inline
+bool ice_ternary_match(const u8 *key, const u8 *key_inv,
+ const u8 *pat, int len)
+{
+ int i;
+
+ for (i = 0; i < len; i++)
+ if (!ice_ternary_match_byte(key[i], key_inv[i], pat[i]))
+ return false;
+
+ return true;
+}
+
+#endif /* _ICE_TMATCH_H_ */
--
2.26.2
[dpdk-dev] [PATCH v3 13/20] net/ice/base: add helper functions for parse graph key matching
Add below two internal helper functions for parse graph key matching
in cam table:
ice_pg_cam_match
ice_pg_nm_cam_match
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_pg_cam.c | 76 +++
drivers/net/ice/base/ice_pg_cam.h | 6 +++
2 files changed, 82 insertions(+)
diff --git a/drivers/net/ice/base/ice_pg_cam.c
b/drivers/net/ice/base/ice_pg_cam.c
index 03484d6a91..fe461ad849 100644
--- a/drivers/net/ice/base/ice_pg_cam.c
+++ b/drivers/net/ice/base/ice_pg_cam.c
@@ -296,3 +296,79 @@ struct ice_pg_nm_cam_item
*ice_pg_nm_sp_cam_table_get(struct ice_hw *hw)
ice_parser_sect_item_get,
_pg_nm_sp_cam_parse_item, false);
}
+
+static bool _pg_cam_match(struct ice_pg_cam_item *item,
+ struct ice_pg_cam_key *key)
+{
+ if (!item->key.valid ||
+ item->key.node_id != key->node_id ||
+ item->key.flag0 != key->flag0 ||
+ item->key.flag1 != key->flag1 ||
+ item->key.flag2 != key->flag2 ||
+ item->key.flag3 != key->flag3 ||
+ item->key.boost_idx != key->boost_idx ||
+ item->key.alu_reg != key->alu_reg ||
+ item->key.next_proto != key->next_proto)
+ return false;
+
+ return true;
+}
+
+static bool _pg_nm_cam_match(struct ice_pg_nm_cam_item *item,
+struct ice_pg_cam_key *key)
+{
+ if (!item->key.valid ||
+ item->key.node_id != key->node_id ||
+ item->key.flag0 != key->flag0 ||
+ item->key.flag1 != key->flag1 ||
+ item->key.flag2 != key->flag2 ||
+ item->key.flag3 != key->flag3 ||
+ item->key.boost_idx != key->boost_idx ||
+ item->key.alu_reg != key->alu_reg)
+ return false;
+
+ return true;
+}
+
+/**
+ * ice_pg_cam_match - search parse graph cam table by key
+ * @table: parse graph cam table to search
+ * @size: cam table size
+ * @key: search key
+ */
+struct ice_pg_cam_item *ice_pg_cam_match(struct ice_pg_cam_item *table,
+int size, struct ice_pg_cam_key *key)
+{
+ int i;
+
+ for (i = 0; i < size; i++) {
+ struct ice_pg_cam_item *item = &table[i];
+
+ if (_pg_cam_match(item, key))
+ return item;
+ }
+
+ return NULL;
+}
+
+/**
+ * ice_pg_nm_cam_match - search parse graph no match cam table by key
+ * @table: parse graph no match cam table to search
+ * @size: cam table size
+ * @key: search key
+ */
+struct ice_pg_nm_cam_item *
+ice_pg_nm_cam_match(struct ice_pg_nm_cam_item *table, int size,
+ struct ice_pg_cam_key *key)
+{
+ int i;
+
+ for (i = 0; i < size; i++) {
+ struct ice_pg_nm_cam_item *item = &table[i];
+
+ if (_pg_nm_cam_match(item, key))
+ return item;
+ }
+
+ return NULL;
+}
diff --git a/drivers/net/ice/base/ice_pg_cam.h
b/drivers/net/ice/base/ice_pg_cam.h
index fcb2e11e54..aeadc20a77 100644
--- a/drivers/net/ice/base/ice_pg_cam.h
+++ b/drivers/net/ice/base/ice_pg_cam.h
@@ -65,4 +65,10 @@ struct ice_pg_cam_item *ice_pg_sp_cam_table_get(struct
ice_hw *hw);
struct ice_pg_nm_cam_item *ice_pg_nm_cam_table_get(struct ice_hw *hw);
struct ice_pg_nm_cam_item *ice_pg_nm_sp_cam_table_get(struct ice_hw *hw);
+
+struct ice_pg_cam_item *ice_pg_cam_match(struct ice_pg_cam_item *table,
+int size, struct ice_pg_cam_key *key);
+struct ice_pg_nm_cam_item *
+ice_pg_nm_cam_match(struct ice_pg_nm_cam_item *table, int size,
+ struct ice_pg_cam_key *key);
#endif /* _ICE_PG_CAM_H_ */
--
2.26.2
[dpdk-dev] [PATCH v3 14/20] net/ice/base: add helper function for ptype markers match
Add internal helper function ice_ptype_mk_tcam_match for ptype markers
matching in tcam table.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_ptype_mk.c | 22 ++
drivers/net/ice/base/ice_ptype_mk.h | 3 +++
2 files changed, 25 insertions(+)
diff --git a/drivers/net/ice/base/ice_ptype_mk.c
b/drivers/net/ice/base/ice_ptype_mk.c
index 33623dcfbc..97c41cb586 100644
--- a/drivers/net/ice/base/ice_ptype_mk.c
+++ b/drivers/net/ice/base/ice_ptype_mk.c
@@ -52,3 +52,25 @@ struct ice_ptype_mk_tcam_item
*ice_ptype_mk_tcam_table_get(struct ice_hw *hw)
ice_parser_sect_item_get,
_parse_ptype_mk_tcam_item, true);
}
+
+/**
+ * ice_ptype_mk_tcam_match - match a pattern on a ptype marker tcam table
+ * @table: ptype marker tcam table to search
+ * @pat: pattern to match
+ * @len: length of the pattern
+ */
+struct ice_ptype_mk_tcam_item *
+ice_ptype_mk_tcam_match(struct ice_ptype_mk_tcam_item *table,
+ u8 *pat, int len)
+{
+ int i;
+
+ for (i = 0; i < ICE_PTYPE_MK_TCAM_TABLE_SIZE; i++) {
+ struct ice_ptype_mk_tcam_item *item = &table[i];
+
+ if (ice_ternary_match(item->key, item->key_inv, pat, len))
+ return item;
+ }
+
+ return NULL;
+}
diff --git a/drivers/net/ice/base/ice_ptype_mk.h
b/drivers/net/ice/base/ice_ptype_mk.h
index c93cd8f0ad..2cd49b1b63 100644
--- a/drivers/net/ice/base/ice_ptype_mk.h
+++ b/drivers/net/ice/base/ice_ptype_mk.h
@@ -15,4 +15,7 @@ struct ice_ptype_mk_tcam_item {
void ice_ptype_mk_tcam_dump(struct ice_hw *hw,
struct ice_ptype_mk_tcam_item *item);
struct ice_ptype_mk_tcam_item *ice_ptype_mk_tcam_table_get(struct ice_hw *hw);
+struct ice_ptype_mk_tcam_item *
+ice_ptype_mk_tcam_match(struct ice_ptype_mk_tcam_item *table,
+ u8 *pat, int len);
#endif /* _ICE_PTYPE_MK_H_ */
--
2.26.2
[dpdk-dev] [PATCH v3 15/20] net/ice/base: add helper function to redirect flags
Add internal helper function ice_flg_redirect to redirect parser flags
to packet flags.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_flg_rd.c | 23 +++
drivers/net/ice/base/ice_flg_rd.h | 1 +
2 files changed, 24 insertions(+)
diff --git a/drivers/net/ice/base/ice_flg_rd.c
b/drivers/net/ice/base/ice_flg_rd.c
index 292916d9a8..833986cac3 100644
--- a/drivers/net/ice/base/ice_flg_rd.c
+++ b/drivers/net/ice/base/ice_flg_rd.c
@@ -51,3 +51,26 @@ struct ice_flg_rd_item *ice_flg_rd_table_get(struct ice_hw
*hw)
ice_parser_sect_item_get,
_flg_rd_parse_item, false);
}
+
+/**
+ * ice_flg_redirect - redirect a parser flag to packet flag
+ * @table: flag redirect table
+ * @psr_flg: parser flag to redirect
+ */
+u64 ice_flg_redirect(struct ice_flg_rd_item *table, u64 psr_flg)
+{
+ u64 flg = 0;
+ int i;
+
+ for (i = 0; i < 64; i++) {
+ struct ice_flg_rd_item *item = &table[i];
+
+ if (!item->expose)
+ continue;
+
+ if (psr_flg & (1ul << item->intr_flg_id))
+ flg |= (1ul << i);
+ }
+
+ return flg;
+}
diff --git a/drivers/net/ice/base/ice_flg_rd.h
b/drivers/net/ice/base/ice_flg_rd.h
index e65350f18c..6c3e01b0fa 100644
--- a/drivers/net/ice/base/ice_flg_rd.h
+++ b/drivers/net/ice/base/ice_flg_rd.h
@@ -13,4 +13,5 @@ struct ice_flg_rd_item {
void ice_flg_rd_dump(struct ice_hw *hw, struct ice_flg_rd_item *item);
struct ice_flg_rd_item *ice_flg_rd_table_get(struct ice_hw *hw);
+u64 ice_flg_redirect(struct ice_flg_rd_item *table, u64 psr_flg);
#endif /* _ICE_FLG_RD_H_ */
--
2.26.2
[dpdk-dev] [PATCH v3 16/20] net/ice/base: add helper function to aggregate flags
Add internal helper function ice_xlt_kb_flg_get to aggregate 64 bit
packet flag into 16 bit key builder flags.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_xlt_kb.c | 27 +++
drivers/net/ice/base/ice_xlt_kb.h | 1 +
2 files changed, 28 insertions(+)
diff --git a/drivers/net/ice/base/ice_xlt_kb.c
b/drivers/net/ice/base/ice_xlt_kb.c
index 8b4043a836..4c1ab747cf 100644
--- a/drivers/net/ice/base/ice_xlt_kb.c
+++ b/drivers/net/ice/base/ice_xlt_kb.c
@@ -187,3 +187,30 @@ struct ice_xlt_kb *ice_xlt_kb_get_rss(struct ice_hw *hw)
{
return _xlt_kb_get(hw, ICE_SID_XLT_KEY_BUILDER_RSS);
}
+
+/**
+ * ice_xlt_kb_flag_get - aggregate 64 bits packet flag into 16 bits xlt flag
+ * @kb: xlt key build
+ * @pkt_flag: 64 bits packet flag
+ */
+u16 ice_xlt_kb_flag_get(struct ice_xlt_kb *kb, u64 pkt_flag)
+{
+ struct ice_xlt_kb_entry *entry = &kb->entries[0];
+ u16 flg = 0;
+ int i;
+
+ /* check flag 15 */
+ if (kb->flag15 & pkt_flag)
+ flg = (u16)(1u << 15);
+
+ /* check flag 0 - 14 */
+ for (i = 0; i < 15; i++) {
+ /* only check first entry */
+ u16 idx = (u16)(entry->flg0_14_sel[i] & 0x3f);
+
+ if (pkt_flag & (1ul << idx))
+ flg |= (u16)(1u << i);
+ }
+
+ return flg;
+}
diff --git a/drivers/net/ice/base/ice_xlt_kb.h
b/drivers/net/ice/base/ice_xlt_kb.h
index a95d845f89..ec802b663a 100644
--- a/drivers/net/ice/base/ice_xlt_kb.h
+++ b/drivers/net/ice/base/ice_xlt_kb.h
@@ -30,4 +30,5 @@ struct ice_xlt_kb *ice_xlt_kb_get_sw(struct ice_hw *hw);
struct ice_xlt_kb *ice_xlt_kb_get_acl(struct ice_hw *hw);
struct ice_xlt_kb *ice_xlt_kb_get_fd(struct ice_hw *hw);
struct ice_xlt_kb *ice_xlt_kb_get_rss(struct ice_hw *hw);
+u16 ice_xlt_kb_flag_get(struct ice_xlt_kb *kb, u64 pkt_flag);
#endif /* _ICE_XLT_KB_H */
--
2.26.2
[dpdk-dev] [PATCH v3 17/20] net/ice/base: add parser execution main loop
Implement function ice_parser_rt_execute which perform the main
loop of the parser.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser_rt.c | 782 ++-
drivers/net/ice/base/ice_parser_rt.h | 27 +-
2 files changed, 803 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ice/base/ice_parser_rt.c
b/drivers/net/ice/base/ice_parser_rt.c
index d62d0170e5..7a44675737 100644
--- a/drivers/net/ice/base/ice_parser_rt.c
+++ b/drivers/net/ice/base/ice_parser_rt.c
@@ -34,12 +34,40 @@ static void _rt_nn_set(struct ice_parser_rt *rt, u16 node)
rt->gpr[GPR_NN_IDX] = node;
}
-static void _rt_flag_set(struct ice_parser_rt *rt, int idx)
+static void _rt_flag_set(struct ice_parser_rt *rt, int idx, bool val)
{
int y = idx / 16;
int x = idx % 16;
- rt->gpr[GPR_FLG_IDX + y] |= (u16)(1 << x);
+ if (val)
+ rt->gpr[GPR_FLG_IDX + y] |= (u16)(1 << x);
+ else
+ rt->gpr[GPR_FLG_IDX + y] &= ~(u16)(1 << x);
+
+ ice_debug(rt->psr->hw, ICE_DBG_PARSER, "Set parser flag %d value %d\n",
+ idx, val);
+}
+
+static void _rt_gpr_set(struct ice_parser_rt *rt, int idx, u16 val)
+{
+ if (idx == GPR_HO_IDX)
+ _rt_ho_set(rt, val);
+ else
+ rt->gpr[idx] = val;
+
+ ice_debug(rt->psr->hw, ICE_DBG_PARSER, "Set GPR %d value %d\n",
+ idx, val);
+}
+
+static void _rt_err_set(struct ice_parser_rt *rt, int idx, bool val)
+{
+ if (val)
+ rt->gpr[GPR_ERR_IDX] |= (u16)(1 << idx);
+ else
+ rt->gpr[GPR_ERR_IDX] &= ~(u16)(1 << idx);
+
+ ice_debug(rt->psr->hw, ICE_DBG_PARSER, "Set parser error %d value %d\n",
+ idx, val);
}
/**
@@ -61,7 +89,7 @@ void ice_parser_rt_reset(struct ice_parser_rt *rt)
for (i = 0; i < 64; i++) {
if ((mi->flags & (1ul << i)) != 0ul)
- _rt_flag_set(rt, i);
+ _rt_flag_set(rt, i, true);
}
rt->psr = psr;
@@ -82,8 +110,655 @@ void ice_parser_rt_pktbuf_set(struct ice_parser_rt *rt,
const u8 *pkt_buf,
ice_memcpy(rt->pkt_buf, pkt_buf, len, ICE_NONDMA_TO_NONDMA);
rt->pkt_len = pkt_len;
- ice_memcpy(&rt->gpr[GPR_HB_IDX], &rt->pkt_buf[ho], 32,
+ ice_memcpy(&rt->gpr[GPR_HB_IDX], &rt->pkt_buf[ho],
+ ICE_PARSER_HDR_BUF_LEN, ICE_NONDMA_TO_NONDMA);
+}
+
+static void _bst_key_init(struct ice_parser_rt *rt, struct ice_imem_item *imem)
+{
+ int second_last_key_idx = ICE_PARSER_BST_KEY_LEN - 2;
+ int last_key_idx = ICE_PARSER_BST_KEY_LEN - 1;
+ u8 tsr = (u8)rt->gpr[GPR_TSR_IDX];
+ u16 ho = rt->gpr[GPR_HO_IDX];
+ u8 *key = rt->bst_key;
+
+ int i, j;
+
+ if (imem->b_kb.tsr_ctrl)
+ key[last_key_idx] = (u8)tsr;
+ else
+ key[last_key_idx] = imem->b_kb.priority;
+
+ for (i = second_last_key_idx; i >= 0; i--) {
+ j = ho + second_last_key_idx - i;
+ if (j < ICE_PARSER_MAX_PKT_LEN)
+ key[i] = rt->pkt_buf[ho + second_last_key_idx - i];
+ else
+ key[i] = 0;
+ }
+
+ ice_debug(rt->psr->hw, ICE_DBG_PARSER, "Generated Boost TCAM Key:\n");
+ ice_debug(rt->psr->hw, ICE_DBG_PARSER, "%02X %02X %02X %02X %02X %02X
%02X %02X %02X %02X %02X %02X %02X %02X %02X %02X %02X %02X %02X %02X\n",
+ key[0], key[1], key[2], key[3], key[4],
+ key[5], key[6], key[7], key[8], key[9],
+ key[10], key[11], key[12], key[13], key[14],
+ key[15], key[16], key[17], key[18], key[19]);
+ ice_debug(rt->psr->hw, ICE_DBG_PARSER, "\n");
+}
+
+static u8 _bit_rev_u8(u8 v)
+{
+ u8 r = 0;
+ int i;
+
+ for (i = 0; i < 8; i++) {
+ r |= (u8)((v & 0x1) << (7 - i));
+ v >>= 1;
+ }
+
+ return r;
+}
+
+static u8 _bit_rev_u16(u16 v, int len)
+{
+ u16 r = 0;
+ int i;
+
+ for (i = 0; i < len; i++) {
+ r |= (u16)((v & 0x1) << (len - 1 - i));
+ v >>= 1;
+ }
+
+ return r;
+}
+
+static u32 _bit_rev_u32(u32 v, int len)
+{
+ u32 r = 0;
+ int i;
+
+ for (i = 0; i < len; i++) {
+ r |= (u32)((v & 0x1) << (len - 1 - i));
+ v >>= 1;
+ }
+
+ return r;
+}
+
+static u32 _hv_bit_sel(struct ice_parser_rt *rt, int start, int len)
+{
+ u64 d64, msk;
+ u8 b[8];
+ int i;
+
+ int offset = GPR_HB_IDX + start / 16;
+
+ ice_memcpy(b, &rt->gpr[offset], 8, ICE_NONDMA_TO_NONDMA);
+
+ for (i = 0; i < 8; i++)
+ b[i] = _bit_rev_u8(b[i]);
+
+ d64 = *(u64 *)&b[0];
+ msk = (1ul << len) - 1;
+
+ return _bit_rev_u32((u32)((d64 >> (start % 16)) & msk), len);
+}
+
+static u32 _pk_build(struct ice_parser_rt *rt, struct ice_np_keybuilder *kb)
+{
+
[dpdk-dev] [PATCH v3 18/20] net/ice/base: support double VLAN mode configure for parser
Add API ice_parser_dvm_set to support turn on/off parser's
double vlan mode.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_bst_tcam.c | 28
drivers/net/ice/base/ice_bst_tcam.h | 4
drivers/net/ice/base/ice_parser.c | 27 +++
drivers/net/ice/base/ice_parser.h | 1 +
4 files changed, 60 insertions(+)
diff --git a/drivers/net/ice/base/ice_bst_tcam.c
b/drivers/net/ice/base/ice_bst_tcam.c
index 76b3a5c551..306f62db2a 100644
--- a/drivers/net/ice/base/ice_bst_tcam.c
+++ b/drivers/net/ice/base/ice_bst_tcam.c
@@ -261,3 +261,31 @@ ice_bst_tcam_match(struct ice_bst_tcam_item *tcam_table,
u8 *pat)
return NULL;
}
+
+static bool _start_with(const char *prefix, const char *string)
+{
+ int len1 = strlen(prefix);
+ int len2 = strlen(string);
+
+ if (len2 < len1)
+ return false;
+
+ return !memcmp(prefix, string, len1);
+}
+
+struct ice_bst_tcam_item *
+ice_bst_tcam_search(struct ice_bst_tcam_item *tcam_table,
+ struct ice_lbl_item *lbl_table,
+ const char *prefix, u16 *start)
+{
+ u16 i = *start;
+
+ for (; i < ICE_BST_TCAM_TABLE_SIZE; i++) {
+ if (_start_with(prefix, lbl_table[i].label)) {
+ *start = i;
+ return &tcam_table[lbl_table[i].idx];
+ }
+ }
+
+ return NULL;
+}
diff --git a/drivers/net/ice/base/ice_bst_tcam.h
b/drivers/net/ice/base/ice_bst_tcam.h
index 3cba0bbf55..e4c96c439d 100644
--- a/drivers/net/ice/base/ice_bst_tcam.h
+++ b/drivers/net/ice/base/ice_bst_tcam.h
@@ -28,4 +28,8 @@ struct ice_lbl_item *ice_bst_lbl_table_get(struct ice_hw *hw);
struct ice_bst_tcam_item *
ice_bst_tcam_match(struct ice_bst_tcam_item *tcam_table, u8 *pat);
+struct ice_bst_tcam_item *
+ice_bst_tcam_search(struct ice_bst_tcam_item *tcam_table,
+ struct ice_lbl_item *lbl_table,
+ const char *prefix, u16 *start);
#endif /*_ICE_BST_TCAM_H_ */
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 8f423710fc..395ad220a0 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -340,3 +340,30 @@ void ice_parser_result_dump(struct ice_hw *hw, struct
ice_parser_result *rslt)
ice_info(hw, "flags_fd = 0x%04x\n", rslt->flags_fd);
ice_info(hw, "flags_rss = 0x%04x\n", rslt->flags_rss);
}
+
+static void _bst_vm_set(struct ice_parser *psr, const char *prefix, bool on)
+{
+ struct ice_bst_tcam_item *item;
+ u16 i = 0;
+
+ while (true) {
+ item = ice_bst_tcam_search(psr->bst_tcam_table,
+ psr->bst_lbl_table,
+ prefix, &i);
+ if (!item)
+ break;
+ item->key[0] = (u8)(on ? 0xff : 0xfe);
+ item->key_inv[0] = (u8)(on ? 0xff : 0xfe);
+ i++;
+ }
+}
+
+/**
+ * ice_parser_dvm_set - configure double vlan mode for parser
+ * @psr: pointer to a parser instance
+ */
+void ice_parser_dvm_set(struct ice_parser *psr, bool on)
+{
+ _bst_vm_set(psr, "BOOST_MAC_VLAN_DVM", on);
+ _bst_vm_set(psr, "BOOST_MAC_VLAN_SVM", !on);
+}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index 715158d6eb..e310466678 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -57,6 +57,7 @@ struct ice_parser {
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
void ice_parser_destroy(struct ice_parser *psr);
+void ice_parser_dvm_set(struct ice_parser *psr, bool on);
struct ice_parser_proto_off {
u8 proto_id; /* hardware protocol ID */
--
2.26.2
[dpdk-dev] [PATCH v3 19/20] net/ice/base: add tunnel port support for parser
UDP tunnel can be added/deleted for vxlan, geneve, ecpri through
below APIs:
ice_parser_vxlan_tunnel_set
ice_parser_geneve_tunnel_set
ice_parser_ecpri_tunnel_set
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser.c | 75 +++
drivers/net/ice/base/ice_parser.h | 6 +++
2 files changed, 81 insertions(+)
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index 395ad220a0..a657e826a3 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -367,3 +367,78 @@ void ice_parser_dvm_set(struct ice_parser *psr, bool on)
_bst_vm_set(psr, "BOOST_MAC_VLAN_DVM", on);
_bst_vm_set(psr, "BOOST_MAC_VLAN_SVM", !on);
}
+
+static enum ice_status
+_tunnel_port_set(struct ice_parser *psr, const char *prefix, u16 udp_port,
+bool on)
+{
+ u8 *buf = (u8 *)&udp_port;
+ struct ice_bst_tcam_item *item;
+ u16 i = 0;
+
+ while (true) {
+ item = ice_bst_tcam_search(psr->bst_tcam_table,
+ psr->bst_lbl_table,
+ prefix, &i);
+ if (!item)
+ break;
+
+ /* found empty slot to add */
+ if (on && item->key[16] == 0xfe && item->key_inv[16] == 0xfe) {
+ item->key_inv[15] = buf[0];
+ item->key_inv[16] = buf[1];
+ item->key[15] = (u8)(0xff - buf[0]);
+ item->key[16] = (u8)(0xff - buf[1]);
+
+ return ICE_SUCCESS;
+ /* found a matched slot to delete */
+ } else if (!on && (item->key_inv[15] == buf[0] ||
+ item->key_inv[16] == buf[1])) {
+ item->key_inv[15] = 0xff;
+ item->key_inv[16] = 0xfe;
+ item->key[15] = 0xff;
+ item->key[16] = 0xfe;
+
+ return ICE_SUCCESS;
+ }
+ i++;
+ }
+
+ return ICE_ERR_PARAM;
+}
+
+/**
+ * ice_parser_vxlan_tunnel_set - configure vxlan tunnel for parser
+ * @psr: pointer to a parser instance
+ * @udp_port: vxlan tunnel port in UDP header
+ * @on: true to turn on; false to turn off
+ */
+enum ice_status ice_parser_vxlan_tunnel_set(struct ice_parser *psr,
+ u16 udp_port, bool on)
+{
+ return _tunnel_port_set(psr, "TNL_VXLAN", udp_port, on);
+}
+
+/**
+ * ice_parser_geneve_tunnel_set - configure geneve tunnel for parser
+ * @psr: pointer to a parser instance
+ * @udp_port: geneve tunnel port in UDP header
+ * @on: true to turn on; false to turn off
+ */
+enum ice_status ice_parser_geneve_tunnel_set(struct ice_parser *psr,
+u16 udp_port, bool on)
+{
+ return _tunnel_port_set(psr, "TNL_GENEVE", udp_port, on);
+}
+
+/**
+ * ice_parser_ecpri_tunnel_set - configure ecpri tunnel for parser
+ * @psr: pointer to a parser instance
+ * @udp_port: ecpri tunnel port in UDP header
+ * @on: true to turn on; false to turn off
+ */
+enum ice_status ice_parser_ecpri_tunnel_set(struct ice_parser *psr,
+ u16 udp_port, bool on)
+{
+ return _tunnel_port_set(psr, "TNL_UDP_ECPRI", udp_port, on);
+}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index e310466678..105d908ebe 100644
--- a/drivers/net/ice/base/ice_parser.h
+++ b/drivers/net/ice/base/ice_parser.h
@@ -58,6 +58,12 @@ struct ice_parser {
enum ice_status ice_parser_create(struct ice_hw *hw, struct ice_parser **psr);
void ice_parser_destroy(struct ice_parser *psr);
void ice_parser_dvm_set(struct ice_parser *psr, bool on);
+enum ice_status ice_parser_vxlan_tunnel_set(struct ice_parser *psr,
+ u16 udp_port, bool on);
+enum ice_status ice_parser_geneve_tunnel_set(struct ice_parser *psr,
+u16 udp_port, bool on);
+enum ice_status ice_parser_ecpri_tunnel_set(struct ice_parser *psr,
+ u16 udp_port, bool on);
struct ice_parser_proto_off {
u8 proto_id; /* hardware protocol ID */
--
2.26.2
[dpdk-dev] [PATCH v3 20/20] net/ice/base: add API for parser profile initialization
Add API ice_parser_profile_init to init a parser profile base on
a parser result and a mask buffer. The ice_parser_profile can feed to
low level FXP engine to create HW profile / field vector directly.
Signed-off-by: Qi Zhang
Acked-by: Junfeng Guo
---
drivers/net/ice/base/ice_parser.c | 112 ++
drivers/net/ice/base/ice_parser.h | 24 +++
2 files changed, 136 insertions(+)
diff --git a/drivers/net/ice/base/ice_parser.c
b/drivers/net/ice/base/ice_parser.c
index a657e826a3..690004e6e2 100644
--- a/drivers/net/ice/base/ice_parser.c
+++ b/drivers/net/ice/base/ice_parser.c
@@ -442,3 +442,115 @@ enum ice_status ice_parser_ecpri_tunnel_set(struct
ice_parser *psr,
{
return _tunnel_port_set(psr, "TNL_UDP_ECPRI", udp_port, on);
}
+
+static bool _nearest_proto_id(struct ice_parser_result *rslt, u16 offset,
+ u8 *proto_id, u16 *proto_off)
+{
+ u16 dist = 0x;
+ u8 p = 0;
+ int i;
+
+ for (i = 0; i < rslt->po_num; i++) {
+ if (offset < rslt->po[i].offset)
+ continue;
+ if (offset - rslt->po[i].offset < dist) {
+ p = rslt->po[i].proto_id;
+ dist = offset - rslt->po[i].offset;
+ }
+ }
+
+ if (dist % 2)
+ return false;
+
+ *proto_id = p;
+ *proto_off = dist;
+
+ return true;
+}
+
+/** default flag mask to cover GTP_EH_PDU, GTP_EH_PDU_LINK and TUN2
+ * In future, the flag masks should learn from DDP
+ */
+#define ICE_KEYBUILD_FLAG_MASK_DEFAULT_SW 0x4002
+#define ICE_KEYBUILD_FLAG_MASK_DEFAULT_ACL 0x
+#define ICE_KEYBUILD_FLAG_MASK_DEFAULT_FD 0x6080
+#define ICE_KEYBUILD_FLAG_MASK_DEFAULT_RSS 0x6010
+
+/**
+ * ice_parser_profile_init - initialize a FXP profile base on parser result
+ * @rslt: a instance of a parser result
+ * @pkt_buf: packet data buffer
+ * @pkt_msk: packet mask buffer
+ * @pkt_len: packet length
+ * @blk: FXP pipeline stage
+ * @prefix_match: match protocol stack exactly or only prefix
+ * @prof: input/output parameter to save the profile
+ */
+enum ice_status ice_parser_profile_init(struct ice_parser_result *rslt,
+ const u8 *pkt_buf, const u8 *msk_buf,
+ int buf_len, enum ice_block blk,
+ bool prefix_match,
+ struct ice_parser_profile *prof)
+{
+ u8 proto_id = 0xff;
+ u16 proto_off = 0;
+ u16 off;
+
+ ice_memset(prof, 0, sizeof(*prof), ICE_NONDMA_MEM);
+ ice_set_bit(rslt->ptype, prof->ptypes);
+ if (blk == ICE_BLK_SW) {
+ prof->flags = rslt->flags_sw;
+ prof->flags_msk = ICE_KEYBUILD_FLAG_MASK_DEFAULT_SW;
+ } else if (blk == ICE_BLK_ACL) {
+ prof->flags = rslt->flags_acl;
+ prof->flags_msk = ICE_KEYBUILD_FLAG_MASK_DEFAULT_ACL;
+ } else if (blk == ICE_BLK_FD) {
+ prof->flags = rslt->flags_fd;
+ prof->flags_msk = ICE_KEYBUILD_FLAG_MASK_DEFAULT_FD;
+ } else if (blk == ICE_BLK_RSS) {
+ prof->flags = rslt->flags_rss;
+ prof->flags_msk = ICE_KEYBUILD_FLAG_MASK_DEFAULT_RSS;
+ } else {
+ return ICE_ERR_PARAM;
+ }
+
+ for (off = 0; off < buf_len - 1; off++) {
+ if (msk_buf[off] == 0 && msk_buf[off + 1] == 0)
+ continue;
+ if (!_nearest_proto_id(rslt, off, &proto_id, &proto_off))
+ continue;
+ if (prof->fv_num >= 32)
+ return ICE_ERR_PARAM;
+
+ prof->fv[prof->fv_num].proto_id = proto_id;
+ prof->fv[prof->fv_num].offset = proto_off;
+ prof->fv[prof->fv_num].spec = *(const u16 *)&pkt_buf[off];
+ prof->fv[prof->fv_num].msk = *(const u16 *)&msk_buf[off];
+ prof->fv_num++;
+ }
+
+ return ICE_SUCCESS;
+}
+
+/**
+ * ice_parser_profile_dump - dump an FXP profile info
+ * @hw: pointer to the hardware structure
+ * @prof: profile info to dump
+ */
+void ice_parser_profile_dump(struct ice_hw *hw, struct ice_parser_profile
*prof)
+{
+ u16 i;
+
+ ice_info(hw, "ptypes:\n");
+ for (i = 0; i < ICE_FLOW_PTYPE_MAX; i++)
+ if (ice_is_bit_set(prof->ptypes, i))
+ ice_info(hw, "\t%d\n", i);
+
+ for (i = 0; i < prof->fv_num; i++)
+ ice_info(hw, "proto = %d, offset = %d spec = 0x%04x, mask =
0x%04x\n",
+prof->fv[i].proto_id, prof->fv[i].offset,
+prof->fv[i].spec, prof->fv[i].msk);
+
+ ice_info(hw, "flags = 0x%04x\n", prof->flags);
+ ice_info(hw, "flags_msk = 0x%04x\n", prof->flags_msk);
+}
diff --git a/drivers/net/ice/base/ice_parser.h
b/drivers/net/ice/base/ice_parser.h
index 105d908ebe..816aea782a
Re: [dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
On Tue, Sep 21, 2021 at 02:16:35PM +0100, David Hunt wrote: >On 21/9/2021 1:04 PM, David Hunt wrote: > > On 21/9/2021 12:57 PM, Bruce Richardson wrote: > > On Tue, Sep 21, 2021 at 12:50:14PM +0100, David Hunt wrote: > > If the user requests to use an lcore above 128 using -l, > the eal will exit with "EAL: invalid core list syntax" and > very little else useful information. > THis patch adds some extra information suggesting to use --lcores > so that physical cores above RTE_MAX_LCORE (default 128) can be > used. This is achieved by using the --lcores option by mapping > the logical cores in the application to physical cores. > There is no change in functionalty, just additional messages > suggesting how the --lcores option might be used for the supplied > list of lcores. For example, if "-l 12-16,130,132" is used, we > see the following additional output on the command line: > EAL: Error = One of the 7 cores provided exceeds RTE_MAX_LCORE (128) > EAL: Please use --lcores instead, e.g. > > Minor suggestion: it would be good to clarify how to use lcores and > what is > happening here in the example. Something like: "Please use --lcores > instead, to map lower lcore ids onto higher-numbered cores", could > help the > user understand better what is happening. > > Hi Bruce, how about: > EAL: Please use --lcores to map logical cores onto cores > > RTE_LCORE_MAX ,e.g. --lcores 0@12,1@13,2@14,3@15,4@16,5@130,6@132 > Rgds, > Dave. > >I think this should do it, as it clarifies the mapping: > >EAL: lcore 130 >= RTE_MAX_LCORE (128) >EAL: lcore 132 >= RTE_MAX_LCORE (128) >EAL: to use high physical core ids , please use --lcores to map them to >lcore ids below RTE_LCORE_MAX, e.g. '--lcores >0@12,1@13,2@14,3@15,4@16,5@130,6@132' > Text looks good to me. Again minor nits: I think the continued lines should be indented, and you should probably wrap immediately after the "e.g." rather than in the middle of the parameter set. /Bruce
[dpdk-dev] [PATCH] app: fix buffer overrun
This patch fixes a possible buffer overrun problem in crypto perf test.
Previously when user configured aad size is over 12 bytes the copy of template
aad will cause a buffer overrun.
The problem is fixed by only copy up to 12 bytes of aad template.
Fixes: 761a321acf91 ("event/cnxk: support vectorized Tx event fast path" )
Cc: pbhagavat...@marvell.com
Signed-off-by: Przemyslaw Zegan
---
app/test-crypto-perf/cperf_test_vectors.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/app/test-crypto-perf/cperf_test_vectors.c
b/app/test-crypto-perf/cperf_test_vectors.c
index 0af01ff911..2c7e314ec8 100644
--- a/app/test-crypto-perf/cperf_test_vectors.c
+++ b/app/test-crypto-perf/cperf_test_vectors.c
@@ -548,12 +548,16 @@ cperf_test_vector_get_dummy(struct cperf_options *options)
t_vec->aead_key.data = aead_key;
if (options->aead_aad_sz) {
- t_vec->aad.data = rte_malloc(NULL,
+ t_vec->aad.data = rte_zmalloc(NULL,
options->aead_aad_sz, 16);
if (t_vec->aad.data == NULL) {
rte_free(t_vec);
return NULL;
}
+
+ if(options->aead_aad_sz > 12)
+ options->aead_aad_sz = 12;
+
memcpy(t_vec->aad.data, aad, options->aead_aad_sz);
t_vec->aad.phys_addr =
rte_malloc_virt2iova(t_vec->aad.data);
t_vec->aad.length = options->aead_aad_sz;
--
2.17.1
--
Intel Research and Development Ireland Limited
Registered in Ireland
Registered Office: Collinstown Industrial Park, Leixlip, County Kildare
Registered Number: 308263
This e-mail and any attachments may contain confidential material for the sole
use of the intended recipient(s). Any review or distribution by others is
strictly prohibited. If you are not the intended recipient, please contact the
sender and delete all copies.
Re: [dpdk-dev] [PATCH v3 00/20] ice/base: add parser module
> -Original Message- > From: Zhang, Qi Z > Sent: Tuesday, September 21, 2021 9:20 PM > To: Yang, Qiming > Cc: Guo, Junfeng ; dev@dpdk.org; Zhang, Qi Z > > Subject: [PATCH v3 00/20] ice/base: add parser module > > Add the parser module that can parse on a raw packet then figure out the > low-level metadata to program the hardware packet process pipeline for flow > offloading(Switch/FDIR/RSS). This is the pre-step to enable a > protocol-agnostic > flow offloading solution for ice devices that leverage Intel DDP technology. > > -v3: > 1. fix 32 bit compile issue in patch 2/20 > > -v2: > 1. use inclusive word in patch 2/20 > 2. replace magic number with macro in patch 17/20 3. fix couple typos > > Qi Zhang (20): > net/ice/base: add parser create and destroy skeleton > net/ice/base: init imem table for parser > net/ice/base: init metainit table for parser > net/ice/base: init parse graph cam table for parser > net/ice/base: init boost TCAM table for parser > net/ice/base: init ptype marker TCAM table for parser > net/ice/base: init marker group table for parser > net/ice/base: init protocol group table for parser > net/ice/base: init flag redirect table for parser > net/ice/base: init XLT key builder for parser > net/ice/base: add parser runtime skeleton > net/ice/base: add helper function for boost TCAM match > net/ice/base: add helper functions for parse graph key matching > net/ice/base: add helper function for ptype markers match > net/ice/base: add helper function to redirect flags > net/ice/base: add helper function to aggregate flags > net/ice/base: add parser execution main loop > net/ice/base: support double VLAN mode configure for parser > net/ice/base: add tunnel port support for parser > net/ice/base: add API for parser profile initialization > > drivers/net/ice/base/ice_bst_tcam.c| 291 + > drivers/net/ice/base/ice_bst_tcam.h| 35 + > drivers/net/ice/base/ice_common.h | 1 + > drivers/net/ice/base/ice_flex_pipe.c | 4 +- > drivers/net/ice/base/ice_flex_pipe.h | 8 + > drivers/net/ice/base/ice_flex_type.h | 2 + > drivers/net/ice/base/ice_flg_rd.c | 76 +++ > drivers/net/ice/base/ice_flg_rd.h | 17 + > drivers/net/ice/base/ice_imem.c| 244 +++ > drivers/net/ice/base/ice_imem.h| 109 > drivers/net/ice/base/ice_metainit.c| 143 > drivers/net/ice/base/ice_metainit.h| 46 ++ > drivers/net/ice/base/ice_mk_grp.c | 55 ++ > drivers/net/ice/base/ice_mk_grp.h | 15 + > drivers/net/ice/base/ice_parser.c | 556 > drivers/net/ice/base/ice_parser.h | 113 > drivers/net/ice/base/ice_parser_rt.c | 870 + > drivers/net/ice/base/ice_parser_rt.h | 53 ++ > drivers/net/ice/base/ice_parser_util.h | 36 + > drivers/net/ice/base/ice_pg_cam.c | 374 +++ > drivers/net/ice/base/ice_pg_cam.h | 74 +++ > drivers/net/ice/base/ice_proto_grp.c | 108 +++ > drivers/net/ice/base/ice_proto_grp.h | 23 + > drivers/net/ice/base/ice_ptype_mk.c| 76 +++ > drivers/net/ice/base/ice_ptype_mk.h| 21 + > drivers/net/ice/base/ice_tmatch.h | 44 ++ > drivers/net/ice/base/ice_type.h| 1 + > drivers/net/ice/base/ice_xlt_kb.c | 216 ++ > drivers/net/ice/base/ice_xlt_kb.h | 34 + > drivers/net/ice/base/meson.build | 11 + > 30 files changed, 3654 insertions(+), 2 deletions(-) create mode 100644 > drivers/net/ice/base/ice_bst_tcam.c > create mode 100644 drivers/net/ice/base/ice_bst_tcam.h > create mode 100644 drivers/net/ice/base/ice_flg_rd.c create mode 100644 > drivers/net/ice/base/ice_flg_rd.h create mode 100644 > drivers/net/ice/base/ice_imem.c create mode 100644 > drivers/net/ice/base/ice_imem.h create mode 100644 > drivers/net/ice/base/ice_metainit.c > create mode 100644 drivers/net/ice/base/ice_metainit.h > create mode 100644 drivers/net/ice/base/ice_mk_grp.c create mode > 100644 drivers/net/ice/base/ice_mk_grp.h create mode 100644 > drivers/net/ice/base/ice_parser.c create mode 100644 > drivers/net/ice/base/ice_parser.h create mode 100644 > drivers/net/ice/base/ice_parser_rt.c > create mode 100644 drivers/net/ice/base/ice_parser_rt.h > create mode 100644 drivers/net/ice/base/ice_parser_util.h > create mode 100644 drivers/net/ice/base/ice_pg_cam.c create mode > 100644 drivers/net/ice/base/ice_pg_cam.h create mode 100644 > drivers/net/ice/base/ice_proto_grp.c > create mode 100644 drivers/net/ice/base/ice_proto_grp.h > create mode 100644 drivers/net/ice/base/ice_ptype_mk.c > create mode 100644 drivers/net/ice/base/ice_ptype_mk.h > create mode 100644 drivers/net/ice/base/ice_tmatch.h create mode > 100644 drivers/net/ice/base/ice_xlt_kb.c create mode 100644 > drivers/net/ice/base/ice_xlt_kb.h > > -- > 2.26.2 Applied to dpdk-next-net-intel. Thanks Qi
Re: [dpdk-dev] [PATCH v6 1/3] security: enforce semantics for Tx inline processing
> > Not all net PMD's/HW can parse packet and identify L2 header and > L3 header locations on Tx. This is inline with other Tx offloads > requirements such as L3 checksum, L4 checksum offload, etc, > where mbuf.l2_len, mbuf.l3_len etc, needs to be set for HW to be > able to generate checksum. Since Inline IPSec is also such a Tx > offload, some PMD's at least need mbuf.l2_len to be valid to > find L3 header and perform Outbound IPSec processing. > > Hence, this patch updates documentation to enforce setting > mbuf.l2_len while setting PKT_TX_SEC_OFFLOAD in mbuf.ol_flags > for Inline IPSec Crypto / Protocol offload processing to > work on Tx. > > Signed-off-by: Nithin Dabilpuram > Acked-by: Konstantin Ananyev > Acked-by: Akhil Goyal Title updated as mbuf: enforce semantics for Tx inline IPSec processing Series Applied to dpdk-next-crypto
Re: [dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
On Tue, Sep 21, 2021 at 1:50 PM David Hunt wrote:
> static int
> eal_parse_coremask(const char *coremask, int *cores)
> {
> @@ -839,54 +880,89 @@ eal_parse_service_corelist(const char *corelist)
> static int
> eal_parse_corelist(const char *corelist, int *cores)
> {
> - unsigned count = 0;
> + unsigned int count = 0, k;
> char *end = NULL;
> int min, max;
> int idx;
> + int lcores[RTE_MAX_LCORE];
Static array...
"-l 0-RTE_MAX_LCORE" / "-c 0x1" / "-l
0-(RTE_MAX_LCORE-1),0" crash.
Please set RTE_MAX_LCORE to 4 (or something that is smaller than your
system core count) and run the tests I provided in my previous mail.
--
David Marchand
Re: [dpdk-dev] [PATCH 3/3] net/mlx5: Fix RSS RETA update
Hi, Maxime
Very nice catch, thank you for the patch.
Minor typo in the commit message "ithat" -> "that".
Besides this:
Acked-by: Viacheslav Ovsiienko
> -Original Message-
> From: Maxime Coquelin
> Sent: Friday, September 10, 2021 12:18
> To: dev@dpdk.org; chenbo@intel.com; amore...@redhat.com;
> david.march...@redhat.com; andrew.rybche...@oktetlabs.ru;
> ferruh.yi...@intel.com; Michael Baum ; Slava
> Ovsiienko
> Cc: sta...@dpdk.org; NBU-Contact-N?lio Laranjeiro
> ; Maxime Coquelin
>
> Subject: [PATCH 3/3] net/mlx5: Fix RSS RETA update
>
> This patch fixes RETA updating for entries above 64.
> Without ithat, these entries are never updated as calculated mask value will
> always be 0.
>
> Fixes: 634efbc2c8c0 ("mlx5: support RETA query and update")
> Cc: sta...@dpdk.org
> Cc: nelio.laranje...@6wind.com
>
> Signed-off-by: Maxime Coquelin
> ---
> drivers/net/mlx5/mlx5_rss.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/mlx5/mlx5_rss.c b/drivers/net/mlx5/mlx5_rss.c index
> c32129cdc2..6dc52acee0 100644
> --- a/drivers/net/mlx5/mlx5_rss.c
> +++ b/drivers/net/mlx5/mlx5_rss.c
> @@ -211,7 +211,7 @@ mlx5_dev_rss_reta_update(struct rte_eth_dev *dev,
> for (idx = 0, i = 0; (i != reta_size); ++i) {
> idx = i / RTE_RETA_GROUP_SIZE;
> pos = i % RTE_RETA_GROUP_SIZE;
> - if (((reta_conf[idx].mask >> i) & 0x1) == 0)
> + if (((reta_conf[idx].mask >> pos) & 0x1) == 0)
> continue;
> MLX5_ASSERT(reta_conf[idx].reta[pos] < priv->rxqs_n);
> (*priv->reta_idx)[i] = reta_conf[idx].reta[pos];
> --
> 2.31.1
Re: [dpdk-dev] [PATCH] net/mlx5: fix empty err msg in mlx5_flow_tunnel_validate
Hi, Winxu
Thank you for the patch.
> If the mlx5_flow_tunnel_validate validate the flow tunnel rule failed, the
Typo? "validate" -> "validating" ?
> err_msg is empty in the rte_flow_error.
Sorry, what do you mean "empty"? It is NULL, not pointing to empty string "".
Patch looks OK for me, could you, please fix commit message typos and send v2?
With best regards,
Slava
> -Original Message-
> From: dev On Behalf Of we...@ucloud.cn
> Sent: Monday, August 9, 2021 8:44
> To: Gregory Etelson
> Cc: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH] net/mlx5: fix empty err msg in
> mlx5_flow_tunnel_validate
>
> From: wenxu
>
> If the mlx5_flow_tunnel_validate validate the flow tunnel rule failed, the
> err_msg is empty in the rte_flow_error.
>
> Fixes: 4ec6360de37d ("net/mlx5: implement tunnel offload")
>
> Signed-off-by: wenxu
> ---
> drivers/net/mlx5/mlx5_flow.c | 43 --
> -
> 1 file changed, 20 insertions(+), 23 deletions(-)
>
> diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
> index e63a297..3c5aca0 100644
> --- a/drivers/net/mlx5/mlx5_flow.c
> +++ b/drivers/net/mlx5/mlx5_flow.c
> @@ -9081,30 +9081,31 @@ int mlx5_alloc_tunnel_hub(struct
> mlx5_dev_ctx_shared *sh)
> return err;
> }
>
> -static inline bool
> +static inline int
> mlx5_flow_tunnel_validate(struct rte_eth_dev *dev,
> struct rte_flow_tunnel *tunnel,
> - const char *err_msg)
> + struct rte_flow_error *error)
> {
> - err_msg = NULL;
> if (!is_tunnel_offload_active(dev)) {
> - err_msg = "tunnel offload was not activated";
> - goto out;
> + return rte_flow_error_set(error, EINVAL,
> +
> RTE_FLOW_ERROR_TYPE_ACTION_CONF, NULL,
> + "tunnel offload was not activated");
> } else if (!tunnel) {
> - err_msg = "no application tunnel";
> - goto out;
> + return rte_flow_error_set(error, EINVAL,
> +
> RTE_FLOW_ERROR_TYPE_ACTION_CONF, NULL,
> + "no application tunnel");
> }
>
> switch (tunnel->type) {
> default:
> - err_msg = "unsupported tunnel type";
> - goto out;
> + return rte_flow_error_set(error, EINVAL,
> +
> RTE_FLOW_ERROR_TYPE_ACTION_CONF, NULL,
> + "unsupported tunnel type");
> case RTE_FLOW_ITEM_TYPE_VXLAN:
> break;
> }
>
> -out:
> - return !err_msg;
> + return 0;
> }
>
> static int
> @@ -9116,13 +9117,11 @@ int mlx5_alloc_tunnel_hub(struct
> mlx5_dev_ctx_shared *sh) {
> int ret;
> struct mlx5_flow_tunnel *tunnel;
> - const char *err_msg = NULL;
> - bool verdict = mlx5_flow_tunnel_validate(dev, app_tunnel,
> err_msg);
>
> - if (!verdict)
> - return rte_flow_error_set(error, EINVAL,
> -
> RTE_FLOW_ERROR_TYPE_ACTION_CONF, NULL,
> - err_msg);
> + ret = mlx5_flow_tunnel_validate(dev, app_tunnel, error);
> + if (ret < 0)
> + return ret;
> +
> ret = mlx5_get_flow_tunnel(dev, app_tunnel, &tunnel);
> if (ret < 0) {
> return rte_flow_error_set(error, ret, @@ -9143,13 +9142,11
> @@ int mlx5_alloc_tunnel_hub(struct mlx5_dev_ctx_shared *sh) {
> int ret;
> struct mlx5_flow_tunnel *tunnel;
> - const char *err_msg = NULL;
> - bool verdict = mlx5_flow_tunnel_validate(dev, app_tunnel,
> err_msg);
>
> - if (!verdict)
> - return rte_flow_error_set(error, EINVAL,
> - RTE_FLOW_ERROR_TYPE_HANDLE,
> NULL,
> - err_msg);
> + ret = mlx5_flow_tunnel_validate(dev, app_tunnel, error);
> + if (ret < 0)
> + return ret;
> +
> ret = mlx5_get_flow_tunnel(dev, app_tunnel, &tunnel);
> if (ret < 0) {
> return rte_flow_error_set(error, ret,
> --
> 1.8.3.1
Re: [dpdk-dev] [PATCH v1 1/5] eventdev: rx_adapter: add support to configure event buffer size
Hi Jerin,
Please see the replies inline.
> -Original Message-
> From: Jerin Jacob
> Sent: Monday, September 20, 2021 11:50 AM
> To: Naga Harish K, S V
> Cc: Jerin Jacob ; Jayatheerthan, Jay
> ; dpdk-dev ; Kundapura,
> Ganapati
> Subject: Re: [dpdk-dev] [PATCH v1 1/5] eventdev: rx_adapter: add support
> to configure event buffer size
>
> On Sat, Sep 18, 2021 at 6:41 PM Naga Harish K S V
> wrote:
> >
> > Currently Rx event buffer is static array with a default size of
> > 192(6*BATCH_SIZE).
> >
> > ``rte_event_eth_rx_adapter_create2`` api is added which takes ``struct
> > rte_event_eth_rx_adapter_params`` to configure event buffer size in
> > addition other params . The event buffer is allocated dynamically at
> > run time aligned to BATCH_SIZE + 2*BATCH_SIZE.
> >
> > Signed-off-by: Naga Harish K S V
> > Signed-off-by: Ganapati Kundapura
> > ---
> >
> > +/**
> > + * A structure to hold adapter config params */ struct
> > +rte_event_eth_rx_adapter_params {
> > + uint16_t event_buf_size;
> > + /**< size of event buffer for the adapter */
>
> See below.
>
> > +};
> > +
> > /**
> > *
> > * Callback function invoked by the SW adapter before it continues @@
> > -330,6 +339,40 @@ int rte_event_eth_rx_adapter_create_ext(uint8_t id,
> uint8_t dev_id,
> > rte_event_eth_rx_adapter_conf_cb conf_cb,
> > void *conf_arg);
> >
> > +/**
> > + * Create a new ethernet Rx event adapter with the specified identifier.
> > + * This function allocates Rx adapter event buffer with the size
> > +specified
> > + * in rxa_params aligned to BATCH_SIZE plus (BATCH_SIZE+BATCH_SIZE)
> > +and
> > + * uses an internal configuration function that creates an event port.
>
> This function may use for adding another
> rte_event_eth_rx_adapter_params:: value.
> So semantics of rte_event_eth_rx_adapter_params::event_buf_size you
> can document at in that structure. This function, you can tell it adapter
> creation varint with parameters or so See below.
>
The documentation is updated as as per the review comments.
> > + * This default function reconfigures the event device with an
> > + * additional event port and setups up the event port using the port
> > + config
> > + * parameter passed into this function. In case the application needs
> > + more
> > + * control in configuration of the service, it should use the
> > + * rte_event_eth_rx_adapter_create_ext() version.
> > + *
> > + * @param id
> > + * The identifier of the ethernet Rx event adapter.
> > + *
> > + * @param dev_id
> > + * The identifier of the event device to configure.
> > + *
> > + * @param rxa_params
> > + * Pointer to struct rte_event_eth_rx_adapter_params containing
> > + * size to allocate rx event buffer.
>
> Value NULL is allowed to represent the default values or so.
>
The api is updated to treat NULL pointer for adapter params with default values.
> > + *
> > + * @param port_config
> > + * Argument of type *rte_event_port_conf* that is passed to the
> > +conf_cb
> > + * function.
> > + *
> > + * @return
> > + * - 0: Success
> > + * - <0: Error code on failure
> > + */
> > +__rte_experimental
> > +int rte_event_eth_rx_adapter_create2(uint8_t id, uint8_t dev_id,
> > + struct rte_event_eth_rx_adapter_params *rxa_params,
> > + struct rte_event_port_conf *port_config);
>
> Couple of suggestion on API name and prototype:
> - I think, we can remove 2 version and give more meaningful,name like
> rte_event_eth_rx_adapter_create_with_param() or so
>
> - Keep new parameter as last to have better compatibility i.e
> rte_event_eth_rx_adapter_create_with_param(uint8_t id, uint8_t dev_id,
> struct rte_event_port_conf *port_config, struct
> rte_event_eth_rx_adapter_params *rxa_params)
The function name and parameters are adjusted as suggested.
Regards
Harish
Re: [dpdk-dev] [PATCH v1 3/5] eventdev:rx_adapter:add per queue event buffer configure support
Hi Jerin,
> -Original Message-
> From: Jerin Jacob
> Sent: Monday, September 20, 2021 11:53 AM
> To: Naga Harish K, S V ; Pavan Nikhilesh
> ; mattias.ronnblom
>
> Cc: Jerin Jacob ; Jayatheerthan, Jay
> ; dpdk-dev
> Subject: Re: [dpdk-dev] [PATCH v1 3/5] eventdev:rx_adapter:add per queue
> event buffer configure support
>
> On Sat, Sep 18, 2021 at 6:42 PM Naga Harish K S V
> wrote:
> >
> > To configure per queue event buffer size, applications sets
> > ``rte_event_eth_rx_adapter_params::use_queue_event_buf`` flag as true
> > and is passed to `rte_event_eth_rx_adapter_create2` api.
> >
> > The per queue event buffer size is populated in
> > ``rte_event_eth_rx_adapter_queue_conf::event_buf_size`` and passed to
> > ``rte_event_eth_rx_adapter_queue_add``.
> >
> > Signed-off-by: Naga Harish K S V
>
> Please change the subject to
> eventdev/rx_adapter: ...
>
It is updated in latest patch set.
> rest looks good to me.
>
> > ---
> > .../prog_guide/event_ethernet_rx_adapter.rst | 19 ---
> > lib/eventdev/rte_event_eth_rx_adapter.h | 4
> > 2 files changed, 16 insertions(+), 7 deletions(-)
> >
> > diff --git a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> > b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> > index cbf694c66b..55d09dbcb8 100644
> > --- a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> > +++ b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> > @@ -62,12 +62,14 @@ service function and needs to create an event port
> > for it. The callback is expected to fill the ``struct
> > rte_event_eth_rx_adapter_conf structure`` passed to it.
> >
> > -If the application desires to control the event buffer size, it can
> > use the -``rte_event_eth_rx_adapter_create2()`` api. The event buffer
> > size is -specified using ``struct
> rte_event_eth_rx_adapter_params::event_buf_size``.
> > -The function is passed the event device to be associated with the
> > adapter -and port configuration for the adapter to setup an event port
> > if the -adapter needs to use a service function.
> > +If the application desires to control the event buffer size at
> > +adapter level, it can use the ``rte_event_eth_rx_adapter_create2()``
> > +api. The event buffer size is specified using ``struct
> rte_event_eth_rx_adapter_params::event_buf_size``.
> > +To configure the event buffer size at queue level, the boolean flag
> > +``struct rte_event_eth_rx_adapter_params::use_queue_event_buf``
> need
> > +to be set to true. The function is passed the event device to be
> > +associated with the adapter and port configuration for the adapter to
> > +setup an event port if the adapter needs to use a service function.
> >
> > Adding Rx Queues to the Adapter Instance
> >
> > @@ -79,7 +81,9 @@ parameter. Event information for packets from this
> > Rx queue is encoded in the ``ev`` field of ``struct
> > rte_event_eth_rx_adapter_queue_conf``. The servicing_weight member
> of
> > the struct rte_event_eth_rx_adapter_queue_conf
> > is the relative polling frequency of the Rx queue and is applicable
> > when the -adapter uses a service core function.
> > +adapter uses a service core function. The applications can configure
> > +queue event buffer size in ``struct
> > +rte_event_eth_rx_adapter_queue_conf::event_buf_size``
> > +parameter.
> >
> > .. code-block:: c
> >
> > @@ -90,6 +94,7 @@ adapter uses a service core function.
> > queue_config.rx_queue_flags = 0;
> > queue_config.ev = ev;
> > queue_config.servicing_weight = 1;
> > + queue_config.event_buf_size = 1024;
> >
> > err = rte_event_eth_rx_adapter_queue_add(id,
> > eth_dev_id, diff
> > --git a/lib/eventdev/rte_event_eth_rx_adapter.h
> > b/lib/eventdev/rte_event_eth_rx_adapter.h
> > index a1b5e0ed37..f9e63dc126 100644
> > --- a/lib/eventdev/rte_event_eth_rx_adapter.h
> > +++ b/lib/eventdev/rte_event_eth_rx_adapter.h
> > @@ -199,6 +199,8 @@ struct rte_event_eth_rx_adapter_queue_conf {
> > * Valid when RTE_EVENT_ETH_RX_ADAPTER_QUEUE_EVENT_VECTOR
> flag is set in
> > * @see rte_event_eth_rx_adapter_queue_conf::rx_queue_flags.
> > */
> > + uint16_t event_buf_size;
> > + /**< event buffer size for this queue */
> > };
> >
> > /**
> > @@ -263,6 +265,8 @@ struct rte_event_eth_rx_adapter_vector_limits {
> > struct rte_event_eth_rx_adapter_params {
> > uint16_t event_buf_size;
> > /**< size of event buffer for the adapter */
> > + bool use_queue_event_buf;
> > + /**< flag to indicate that event buffer is separate for each
> > + queue */
> > };
> >
> > /**
> > --
> > 2.25.1
> >
Regards
Harish
Re: [dpdk-dev] [PATCH] kni: Fix request overwritten
On 9/17/2021 1:31 PM, Elad Nachman wrote:
> Fix lack of multiple KNI requests handling support by introducing a
> rotating ring mechanism for the sync buffer.
>
Thanks Elad for the patch.
I have missed that 'rte_kni_handle_request()' can be called by multiple cores,
at least kni sample app does that.
In that case having multiple requests may end up multiple control command run at
the same time from different cores and this may lead undefined behavior.
Forcing to have single request at a time also simplifies to seriliaze the
control commands to the device.
What about your suggestion to return -EAGAIN, if the request is not processed by
the userspace. And this can be detected by a new field in 'rte_kni_request'
which kernel sets when the request issued and userspace unset when the
processing done. 'kni_net_process_request()' can require that field to be unset
before putting a new request. What do you think?
> Prevent kni_net_change_rx_flags() from calling kni_net_process_request()
> with a malformed request.
>
This is completely something else so can be dropped from this patch, but OK to
add 'if (req.req_id)' check although current request is harmless since 'req' is
already memset.
> Bugzilla ID: 809
>
> Signed-off-by: Elad Nachman
> ---
> kernel/linux/kni/kni_dev.h | 2 ++
> kernel/linux/kni/kni_misc.c | 2 ++
> kernel/linux/kni/kni_net.c | 25 +
> lib/kni/rte_kni.c | 14 +++---
> lib/kni/rte_kni_common.h| 1 +
> 5 files changed, 29 insertions(+), 15 deletions(-)
>
> diff --git a/kernel/linux/kni/kni_dev.h b/kernel/linux/kni/kni_dev.h
> index c15da311ba..b9e8b3d10d 100644
> --- a/kernel/linux/kni/kni_dev.h
> +++ b/kernel/linux/kni/kni_dev.h
> @@ -74,6 +74,8 @@ struct kni_dev {
>
> void *sync_kva;
> void *sync_va;
> + unsigned int sync_ring_size;
> + atomic_t sync_ring_idx;
>
> void *mbuf_kva;
> void *mbuf_va;
> diff --git a/kernel/linux/kni/kni_misc.c b/kernel/linux/kni/kni_misc.c
> index 2b464c4381..f3cee97818 100644
> --- a/kernel/linux/kni/kni_misc.c
> +++ b/kernel/linux/kni/kni_misc.c
> @@ -345,6 +345,8 @@ kni_ioctl_create(struct net *net, uint32_t ioctl_num,
>
> kni->net_dev = net_dev;
> kni->core_id = dev_info.core_id;
> + kni->sync_ring_size = dev_info.sync_ring_size;
> + kni->sync_ring_idx.counter = 0;
> strncpy(kni->name, dev_info.name, RTE_KNI_NAMESIZE);
>
> /* Translate user space info into kernel space info */
> diff --git a/kernel/linux/kni/kni_net.c b/kernel/linux/kni/kni_net.c
> index 611719b5ee..dc066cdd98 100644
> --- a/kernel/linux/kni/kni_net.c
> +++ b/kernel/linux/kni/kni_net.c
> @@ -110,6 +110,8 @@ kni_net_process_request(struct net_device *dev, struct
> rte_kni_request *req)
> void *resp_va;
> uint32_t num;
> int ret_val;
> + unsigned int idx;
> + char *k_reqptr, *v_reqptr;
>
> ASSERT_RTNL();
>
> @@ -122,10 +124,17 @@ kni_net_process_request(struct net_device *dev, struct
> rte_kni_request *req)
> }
>
> mutex_lock(&kni->sync_lock);
> -
> + idx = atomic_read(&kni->sync_ring_idx);
> + atomic_cmpxchg(&kni->sync_ring_idx, idx,
> + (idx + 1) % kni->sync_ring_size);
> /* Construct data */
> - memcpy(kni->sync_kva, req, sizeof(struct rte_kni_request));
> - num = kni_fifo_put(kni->req_q, &kni->sync_va, 1);
> + k_reqptr = (char *)((uintptr_t)kni->sync_kva +
> + sizeof(struct rte_kni_request) * idx);
> + v_reqptr = (char *)((uintptr_t)kni->sync_va +
> + sizeof(struct rte_kni_request) * idx);
> +
> + memcpy(k_reqptr, req, sizeof(struct rte_kni_request));
> + num = kni_fifo_put(kni->req_q, (void **)&v_reqptr, 1);
> if (num < 1) {
> pr_err("Cannot send to req_q\n");
> ret = -EBUSY;
> @@ -147,14 +156,14 @@ kni_net_process_request(struct net_device *dev, struct
> rte_kni_request *req)
> goto fail;
> }
> num = kni_fifo_get(kni->resp_q, (void **)&resp_va, 1);
> - if (num != 1 || resp_va != kni->sync_va) {
> + if (num != 1) {
> /* This should never happen */
> pr_err("No data in resp_q\n");
> ret = -ENODATA;
> goto fail;
> }
>
> - memcpy(req, kni->sync_kva, sizeof(struct rte_kni_request));
> + memcpy(req, k_reqptr, sizeof(struct rte_kni_request));
> async:
> ret = 0;
>
> @@ -681,13 +690,12 @@ kni_net_change_mtu(struct net_device *dev, int new_mtu)
> static void
> kni_net_change_rx_flags(struct net_device *netdev, int flags)
> {
> - struct rte_kni_request req;
> + struct rte_kni_request req = { 0 };
>
> memset(&req, 0, sizeof(req));
>
> if (flags & IFF_ALLMULTI) {
> req.req_id = RTE_KNI_REQ_CHANGE_ALLMULTI;
> -
> if (netdev->flags & IFF_ALLMULTI)
> req.allmulti = 1;
> else
> @@ -703,7 +711,8 @
Re: [dpdk-dev] [PATCH v3 2/8] dmadev: add burst capacity API
Hi Jerin, > > > > > From: Kevin Laatz <[2]kevin.la...@intel.com> > > > > > Add a burst capacity check API to the dmadev library. This API is > > > > > useful to > > > > > applications which need to how many descriptors can be enqueued > in > > > > > the > > > > > current batch. For example, it could be used to determine whether > > > > > all > > > > > segments of a multi-segment packet can be enqueued in the > > > > > same > > > batch > > > > > or not > > > > > (to avoid half-offload of the packet). > > > > > > > > > > #Could you share more details on the use case with vhost? > > > > ># Are they planning to use this in fast path if so it need to move > > > > > as > > > > >fast path function pointer? > > > > > > > > I believe the intent is to use it on fastpath, but I would assume > > > > only once per burst, so the penalty for non-fastpath may be > > > > acceptable. As you point out - for an app that really doesn't want > > > > to have to pay that penalty, tracking ring use itself is possible. > > > > > > > > The desire for fast-path use is also why I suggested having the > > > > space as an optional return parameter from the submit API call. It > > > > could logically also be a return value from the "completed" call, > > > > which might actually make more sense. > > > > > > > > ># Assume the use case needs N rte_dma_copy to complete a > > > > > logical > > > copy > > > > >at vhost level. Is the any issue in half-offload, meaning when N th > one > > > > >successfully completed then only the logical copy is completed. > Right? > > > > > > > > Yes, as I understand it, the issue is for multi-segment packets, > > > > where we only want to enqueue the first segment if we know we will > > > > success with the final one too. > > > > > > Sorry for the delay in reply. > > > > > > If so, why do we need this API. We can mark a logical transaction > > > completed IFF final segment is succeeded. Since this fastpath API, I > > > would like to really understand the real use case for it, so if > > > required then we need to implement in an optimized way. > > > Otherwise driver does not need to implement this to have generic > > > solution for all the drivers. > > > Hi Jiayu, Sunil, > > > The fact is that it's very hard for apps to calculate the available space > > of a > DMA ring. > > Yes, I agree. > > My question is more why to calculate the space per burst and introduce yet > another fastpath API. > For example, the application needs to copy 8 segments to complete a logical > copy in the application perspective. > In case, when 8th copy is completed then only the application marks the > logical copy completed. > i.e why to check per burst, 8 segments are available or not? Even it is > available, there may be multiple reasons why any of the segment copies can > fail. So the application needs to track all the jobs completed or not anyway. > Am I missing something in terms of vhost or OVS usage? > For the packets that do not entirely fit in the DMA ring , we have a SW copy fallback in place. So, we would like to avoid scenario caused because of DMA ring full where few parts of the packet are copied through DMA and other parts by CPU. Besides, this API would also help improve debuggability/device introspection to check the occupancy rather than the app having to manually track the state of every DMA device in use. Copying from other thread: > What are those scenarios, could you share some descriptions of them. > What if the final or any segment fails event the space is available. > So you have to take care of that anyway. RIght? I think this is app dependent no? The application can choose not to take care of such scenarios and treat the packets as dropped. Ring full scenarios(-ENOSPC from rte_dma_copy) could be avoided with this API but other errors mean a failure which unfortunately cannot be avoided. > > > For DSA, the available space is decided by three factors: the number > > of available slots in SW ring, the max batching size of a batch > > descriptor, and if there are available batch descriptors. The first > > one is configured by SW, and apps can calculate it. But the second depends > on DSA HW, and the third one is hided in DSA driver which is not visible to > apps. > > Considering the complexity of different DMA HW, I think the best way > > is to hide all details inside DMA dev and provide this check capacity API > > for > apps. > > > > Thanks, > > Jiayu > > > > > > > > > > > > > ># There is already nb_desc with which a dma_queue is configured. > So if > > > > >the application does its accounting properly, it knows how many > desc it > > > > >has used up and how many completions it has processed. > > > > > > > > Agreed. It's just more work for the app, and for simplicity and > > > > completeness I think we should add this API. Because there are > > > > other options I think it should be available, but n
Re: [dpdk-dev] [PATCH v2] update Windows roadmap
11/09/2021 01:07, Dmitry Kozlyuk: > Add the current status and the actual roadmap, remove completed entries. > More detailed plan in the mailing list: > https://mails.dpdk.org/archives/dev/2021-August/217463.html > > Signed-off-by: Dmitry Kozlyuk > --- > v2: improve wording, fix typos, remove improbable items (Thomas). Applied, thanks. Note: fixed capitalization VMWare -> VMware
Re: [dpdk-dev] [EXT] [PATCH] app: fix buffer overrun
>This patch fixes a possible buffer overrun problem in crypto perf test.
>Previously when user configured aad size is over 12 bytes the copy of
>template aad will cause a buffer overrun.
>The problem is fixed by only copy up to 12 bytes of aad template.
>
>Fixes: 761a321acf91 ("event/cnxk: support vectorized Tx event fast
>path" )
>Cc: pbhagavat...@marvell.com
>
I don't see how the above mentioned patch is related to test-crypto-perf.
Could you please recheck?
>Signed-off-by: Przemyslaw Zegan
>---
> app/test-crypto-perf/cperf_test_vectors.c | 6 +-
> 1 file changed, 5 insertions(+), 1 deletion(-)
>
>diff --git a/app/test-crypto-perf/cperf_test_vectors.c b/app/test-
>crypto-perf/cperf_test_vectors.c
>index 0af01ff911..2c7e314ec8 100644
>--- a/app/test-crypto-perf/cperf_test_vectors.c
>+++ b/app/test-crypto-perf/cperf_test_vectors.c
>@@ -548,12 +548,16 @@ cperf_test_vector_get_dummy(struct
>cperf_options *options)
> t_vec->aead_key.data = aead_key;
>
> if (options->aead_aad_sz) {
>- t_vec->aad.data = rte_malloc(NULL,
>+ t_vec->aad.data = rte_zmalloc(NULL,
> options->aead_aad_sz, 16);
> if (t_vec->aad.data == NULL) {
> rte_free(t_vec);
> return NULL;
> }
>+
>+ if(options->aead_aad_sz > 12)
>+ options->aead_aad_sz = 12;
>+
> memcpy(t_vec->aad.data, aad, options-
>>aead_aad_sz);
> t_vec->aad.phys_addr =
>rte_malloc_virt2iova(t_vec->aad.data);
> t_vec->aad.length = options->aead_aad_sz;
>--
>2.17.1
>
>--
>Intel Research and Development Ireland Limited
>Registered in Ireland
>Registered Office: Collinstown Industrial Park, Leixlip, County Kildare
>Registered Number: 308263
>
>
>This e-mail and any attachments may contain confidential material for
>the sole
>use of the intended recipient(s). Any review or distribution by others is
>strictly prohibited. If you are not the intended recipient, please contact
>the
>sender and delete all copies.
Re: [dpdk-dev] [PATCH v3 2/8] dmadev: add burst capacity API
On Tue, Sep 21, 2021 at 7:46 PM Pai G, Sunil wrote:
>
> Hi Jerin,
Hi Sunil,
>
> > > > > > From: Kevin Laatz <[2]kevin.la...@intel.com>
> > > > > > Add a burst capacity check API to the dmadev library. This API
> > > > > > is
> > > > > > useful to
> > > > > > applications which need to how many descriptors can be enqueued
> > in
> > > > > > the
> > > > > > current batch. For example, it could be used to determine
> > > > > > whether
> > > > > > all
> > > > > > segments of a multi-segment packet can be enqueued in the
> > > > > > same
> > > > batch
> > > > > > or not
> > > > > > (to avoid half-offload of the packet).
> > > > > >
> > > > > > #Could you share more details on the use case with vhost?
> > > > > ># Are they planning to use this in fast path if so it need to
> > > > > > move as
> > > > > >fast path function pointer?
> > > > >
> > > > > I believe the intent is to use it on fastpath, but I would assume
> > > > > only once per burst, so the penalty for non-fastpath may be
> > > > > acceptable. As you point out - for an app that really doesn't want
> > > > > to have to pay that penalty, tracking ring use itself is possible.
> > > > >
> > > > > The desire for fast-path use is also why I suggested having the
> > > > > space as an optional return parameter from the submit API call. It
> > > > > could logically also be a return value from the "completed" call,
> > > > > which might actually make more sense.
> > > > >
> > > > > ># Assume the use case needs N rte_dma_copy to complete a
> > > > > > logical
> > > > copy
> > > > > >at vhost level. Is the any issue in half-offload, meaning when N
> > > > > > th
> > one
> > > > > >successfully completed then only the logical copy is completed.
> > Right?
> > > > >
> > > > > Yes, as I understand it, the issue is for multi-segment packets,
> > > > > where we only want to enqueue the first segment if we know we will
> > > > > success with the final one too.
> > > >
> > > > Sorry for the delay in reply.
> > > >
> > > > If so, why do we need this API. We can mark a logical transaction
> > > > completed IFF final segment is succeeded. Since this fastpath API, I
> > > > would like to really understand the real use case for it, so if
> > > > required then we need to implement in an optimized way.
> > > > Otherwise driver does not need to implement this to have generic
> > > > solution for all the drivers.
> >
> >
> > Hi Jiayu, Sunil,
> >
> > > The fact is that it's very hard for apps to calculate the available
> > > space of a
> > DMA ring.
> >
> > Yes, I agree.
> >
> > My question is more why to calculate the space per burst and introduce yet
> > another fastpath API.
> > For example, the application needs to copy 8 segments to complete a logical
> > copy in the application perspective.
> > In case, when 8th copy is completed then only the application marks the
> > logical copy completed.
> > i.e why to check per burst, 8 segments are available or not? Even it is
> > available, there may be multiple reasons why any of the segment copies can
> > fail. So the application needs to track all the jobs completed or not
> > anyway.
> > Am I missing something in terms of vhost or OVS usage?
> >
>
> For the packets that do not entirely fit in the DMA ring , we have a SW copy
> fallback in place.
> So, we would like to avoid scenario caused because of DMA ring full where few
> parts of the packet are copied through DMA and other parts by CPU.
> Besides, this API would also help improve debuggability/device introspection
> to check the occupancy rather than the app having to manually track the state
> of every DMA device in use.
To understand it better, Could you share more details on feedback
mechanism on your application enqueue
app_enqueue_v1(.., nb_seg)
{
/* Not enough space, Let application handle it by
dropping or resubmitting */
if (rte_dmadev_burst_capacity() < nb_seg)
return -ENOSPC;
do rte_dma_op() in loop without checking error;
return 0; // Success
}
vs
app_enqueue_v2(.., nb_seg)
{
int rc;
rc |= rte_dma_op() in loop without checking error;
return rc; // return the actual status to application if
Not enough space, Let application handle it by dropping or
resubmitting */
}
Is app_enqueue_v1() and app_enqueue_v2() logically the same from
application PoV. Right?
If not, could you explain, the version you are planning to do for app_enqueue()
> Copying from other thread:
>
> > What are those scenarios, could you share some descriptions of them.
> > What if the final or any segment fails event the space is available.
> > So you have to take care of that anyway. RIght?
>
> I think this is app dependent no? The application can choose not to take
> care of such scenarios and treat the packets as dropped.
> Ring full scenarios(-ENOSPC from rte_dma_copy) could be avoided wi
Re: [dpdk-dev] [PATCH v3 1/2] eal: add additional info if core list too long
On 21/9/2021 2:51 PM, David Marchand wrote:
On Tue, Sep 21, 2021 at 1:50 PM David Hunt wrote:
static int
eal_parse_coremask(const char *coremask, int *cores)
{
@@ -839,54 +880,89 @@ eal_parse_service_corelist(const char *corelist)
static int
eal_parse_corelist(const char *corelist, int *cores)
{
- unsigned count = 0;
+ unsigned int count = 0, k;
char *end = NULL;
int min, max;
int idx;
+ int lcores[RTE_MAX_LCORE];
Static array...
"-l 0-RTE_MAX_LCORE" / "-c 0x1" / "-l
0-(RTE_MAX_LCORE-1),0" crash.
Please set RTE_MAX_LCORE to 4 (or something that is smaller than your
system core count) and run the tests I provided in my previous mail.
Hi David,
I did set the lcore_max to 4, and ran the tests you provided, and they
all looked OK.
However, as you have pointed out, there is a problem with the lcores
array, which I have now fixed, I'll push up shortly. I'm not expecting
to populate with more than RTE_MAX_LCORE elements, but there was an edge
case where one extra was being added.
Thanks again,
Dave.
Re: [dpdk-dev] [PATCH] net/sfc: fix getting accumulative SW xstat
On 9/15/2021 11:40 AM, Andrew Rybchenko wrote:
> From: Ivan Ilchenko
>
> Add missing initialisation of the accumulative SW xstat to
> zero since it is sum of per-queue xstats.
>
> Fixes: fdd7719eb3c ("net/sfc: add xstats for Rx/Tx doorbells")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Ivan Ilchenko
> Signed-off-by: Andrew Rybchenko
Applied to dpdk-next-net/main, thanks.
Re: [dpdk-dev] [PATCH v3 2/8] dmadev: add burst capacity API
Hi Jerin,
> > > > The fact is that it's very hard for apps to calculate the
> > > > available space of a
> > > DMA ring.
> > >
> > > Yes, I agree.
> > >
> > > My question is more why to calculate the space per burst and
> > > introduce yet another fastpath API.
> > > For example, the application needs to copy 8 segments to complete a
> > > logical copy in the application perspective.
> > > In case, when 8th copy is completed then only the application marks
> > > the logical copy completed.
> > > i.e why to check per burst, 8 segments are available or not? Even it
> > > is available, there may be multiple reasons why any of the segment
> > > copies can fail. So the application needs to track all the jobs completed
> > > or
> not anyway.
> > > Am I missing something in terms of vhost or OVS usage?
> > >
> >
> > For the packets that do not entirely fit in the DMA ring , we have a SW copy
> fallback in place.
> > So, we would like to avoid scenario caused because of DMA ring full where
> few parts of the packet are copied through DMA and other parts by CPU.
> > Besides, this API would also help improve debuggability/device
> introspection to check the occupancy rather than the app having to manually
> track the state of every DMA device in use.
>
> To understand it better, Could you share more details on feedback
> mechanism on your application enqueue
>
> app_enqueue_v1(.., nb_seg)
> {
> /* Not enough space, Let application handle it by dropping or
> resubmitting */
> if (rte_dmadev_burst_capacity() < nb_seg)
> return -ENOSPC;
>
> do rte_dma_op() in loop without checking error;
> return 0; // Success
> }
>
> vs
> app_enqueue_v2(.., nb_seg)
> {
>int rc;
>
> rc |= rte_dma_op() in loop without checking error;
> return rc; // return the actual status to application if Not
> enough space,
> Let application handle it by dropping or resubmitting */ }
>
> Is app_enqueue_v1() and app_enqueue_v2() logically the same from
> application PoV. Right?
>
> If not, could you explain, the version you are planning to do for
> app_enqueue()
The exact version can be found here :
http://patchwork.ozlabs.org/project/openvswitch/patch/20210907120021.4093604-2-sunil.pa...@intel.com/
Unfortunately, both versions are not same in our case because of the SW
fallback we have for ring full scenario's.
For a packet with 8 nb_segs, if the ring has only space for 4 , we would avoid
this packet with app_enqueue_v1
while going ahead with an enqueue with app_enqueue_v2, resulting in a mix of
DMA and CPU copies for a packet which we would want to avoid.
Thanks and regards,
Sunil
Re: [dpdk-dev] [PATCH v4 1/5] test/crypto: add lookaside IPsec tests
Hi Anoob,
Few minor comments, Please see inline.
Apart from that,
Acked-by: Akhil Goyal
>
Update title as
Test/crypto: add lookaside IPsec cases.
> +static int
> +security_proto_supported(enum rte_security_session_action_type action,
> + enum rte_security_session_protocol proto);
> +
> +static int
> +dev_configure_and_start(uint64_t ff_disable);
> +
Do we really need to forward declare?
> static struct rte_mbuf *
> setup_test_string(struct rte_mempool *mpool,
> const char *string, size_t len, uint8_t blocksize)
> @@ -753,6 +763,43 @@ crypto_gen_testsuite_setup(void)
>
> #ifdef RTE_LIB_SECURITY
> static int
> +ipsec_proto_testsuite_setup(void)
> +{
> + struct crypto_testsuite_params *ts_params = &testsuite_params;
> + struct crypto_unittest_params *ut_params = &unittest_params;
> + struct rte_cryptodev_info dev_info;
> + int ret = 0;
> +
> + rte_cryptodev_info_get(ts_params->valid_devs[0], &dev_info);
> +
> + if (!(dev_info.feature_flags & RTE_CRYPTODEV_FF_SECURITY)) {
> + RTE_LOG(INFO, USER1, "Feature flag requirements for IPsec
> Proto "
> + "testsuite not met\n");
> + return TEST_SKIPPED;
> + }
> +
> + /* Reconfigure to enable security */
Update comment like
/*Reconfigure to enable security and disable crypto */
BTW, shouldn't this be dev_configure_and_start(0)
Why is sym and asym disabled here?
> + dev_configure_and_start(RTE_CRYPTODEV_FF_SYMMETRIC_CRYPTO
> |
> + RTE_CRYPTODEV_FF_ASYMMETRIC_CRYPTO);
Return value not taken care here.
> +
> + /* Set action type */
> + ut_params->type =
> RTE_SECURITY_ACTION_TYPE_LOOKASIDE_PROTOCOL;
> +
> + if (security_proto_supported(
> +
> RTE_SECURITY_ACTION_TYPE_LOOKASIDE_PROTOCOL,
> + RTE_SECURITY_PROTOCOL_IPSEC) < 0) {
> + RTE_LOG(INFO, USER1, "Capability requirements for IPsec
> Proto "
> + "test not met\n");
> + ret = TEST_SKIPPED;
> + }
> +
> + /* Stop the device */
> + rte_cryptodev_stop(ts_params->valid_devs[0]);
Add a comment that the device will be started again in ut_setup_security()
> +
> + ret = test_ipsec_post_process(ut_params->ibuf, &td[i],
> + res_d_tmp, silent);
> + if (ret != TEST_SUCCESS)
> + goto crypto_op_free;
> +
> + rte_crypto_op_free(ut_params->op);
> + ut_params->op = NULL;
> +
> + rte_pktmbuf_free(ut_params->ibuf);
> + ut_params->ibuf = NULL;
> + }
> +
> +crypto_op_free:
> + rte_crypto_op_free(ut_params->op);
> + ut_params->op = NULL;
> +
> + rte_pktmbuf_free(ut_params->ibuf);
> + ut_params->ibuf = NULL;
> +
Above four lines are getting executed again in the success cases.
[dpdk-dev] [PATCH v2] stack: fix reload head when pop fails
The previous commit 18effad9cfa7 ("stack: reload head when pop fails")
only changed C11 implementation, not generic implementation.
List head must be loaded right before continue (when failed to find the
new head). Without this, one thread might keep trying and failing to pop
items without ever loading the new correct head.
Fixes: 3340202f5954 ("stack: add lock-free implementation")
Cc: sta...@dpdk.org
Signed-off-by: Julien Meunier
---
v2:
* rebase
* update commit log + remove invalid CC email address
lib/stack/rte_stack_lf_generic.h | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/lib/stack/rte_stack_lf_generic.h b/lib/stack/rte_stack_lf_generic.h
index 4850a05ee7..7fa29cedb2 100644
--- a/lib/stack/rte_stack_lf_generic.h
+++ b/lib/stack/rte_stack_lf_generic.h
@@ -128,8 +128,10 @@ __rte_stack_lf_pop_elems(struct rte_stack_lf_list *list,
/* If NULL was encountered, the list was modified while
* traversing it. Retry.
*/
- if (i != num)
+ if (i != num) {
+ old_head = list->head;
continue;
+ }
new_head.top = tmp;
new_head.cnt = old_head.cnt + 1;
--
2.17.1
Re: [dpdk-dev] [PATCH v4 2/5] test/crypto: add combined mode tests
> Subject: [PATCH v4 2/5] test/crypto: add combined mode tests Title: test/crypto: add combined mode IPSec cases > In description explain the meaning of combined mode. Also mention that the encap and decap are done one after the other, Hence known test vectors are not required. Apart from that, Acked-by: Akhil Goyal > Add framework to test IPsec features with all supported > combinations of ciphers. > > Signed-off-by: Anoob Joseph > Signed-off-by: Tejasree Kondoj >
Re: [dpdk-dev] [PATCH v4 02/11] dma/ioat: create dmadev instances on PCI probe
On Fri, Sep 17, 2021 at 03:42:18PM +, Conor Walsh wrote:
When a suitable device is found during the PCI probe, create a dmadev
instance for each channel. Internal structures and HW definitions required
for device creation are also included.
Signed-off-by: Conor Walsh
Reviewed-by: Kevin Laatz
---
drivers/dma/ioat/ioat_dmadev.c | 119 ++-
drivers/dma/ioat/ioat_hw_defs.h | 45
drivers/dma/ioat/ioat_internal.h | 24 +++
3 files changed, 186 insertions(+), 2 deletions(-)
diff --git a/drivers/dma/ioat/ioat_dmadev.c b/drivers/dma/ioat/ioat_dmadev.c
index f3491d45b1..b815d30bcf 100644
--- a/drivers/dma/ioat/ioat_dmadev.c
+++ b/drivers/dma/ioat/ioat_dmadev.c
@@ -4,6 +4,7 @@
+/* Destroy a DMA device. */
+static int
+ioat_dmadev_destroy(const char *name)
+{
+ struct rte_dma_dev *dev;
+ struct ioat_dmadev *ioat;
+ int ret;
+
+ if (!name) {
+ IOAT_PMD_ERR("Invalid device name");
+ return -EINVAL;
+ }
+
+ dev = &rte_dma_devices[rte_dma_get_dev_id(name)];
+ if (!dev) {
+ IOAT_PMD_ERR("Invalid device name (%s)", name);
+ return -EINVAL;
+ }
+
I think you need to independently check the return value from
rte_dma_get_dev_id, rather than assuming when it returns an error value the
resultant index location will hold a null pointer.
I will assign rte_dma_get_dev_id(name) to a variable and check it before
use in v5.
+ ioat = dev->dev_private;
+ if (!ioat) {
+ IOAT_PMD_ERR("Error getting dev_private");
+ return -EINVAL;
+ }
+
+ dev->dev_private = NULL;
+ rte_free(ioat->desc_ring);
+
+ ret = rte_dma_pmd_release(name);
The rte_dma_pmd_allocate function reserves memory for the private data, so
the release function should free that memory too. However, you have
assigned private_data to NULL just above, so that probably won't work.
I think I will actually remove the "dev->dev_private = NULL" in v5 as
this is handled by the lib.
Thanks,
Conor.
+ if (ret)
+ IOAT_PMD_DEBUG("Device cleanup failed");
+
+ return 0;
+}
+
Re: [dpdk-dev] [PATCH v4 11/11] devbind: move ioat device ID for ICX to dmadev category
On Fri, Sep 17, 2021 at 03:42:27PM +, Conor Walsh wrote: Move Intel IOAT devices on Ice Lake systems from Misc to DMA devices. Signed-off-by: Conor Walsh Reviewed-by: Kevin Laatz --- usertools/dpdk-devbind.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/usertools/dpdk-devbind.py b/usertools/dpdk-devbind.py index 98b698ccc0..afebc8cb62 100755 --- a/usertools/dpdk-devbind.py +++ b/usertools/dpdk-devbind.py @@ -69,14 +69,13 @@ network_devices = [network_class, cavium_pkx, avp_vnic, ifpga_class] baseband_devices = [acceleration_class] crypto_devices = [encryption_class, intel_processor_class] -dma_devices = [intel_idxd_spr] +dma_devices = [intel_idxd_spr, intel_ioat_icx] eventdev_devices = [cavium_sso, cavium_tim, intel_dlb, octeontx2_sso] mempool_devices = [cavium_fpa, octeontx2_npa] compress_devices = [cavium_zip] regex_devices = [octeontx2_ree] misc_devices = [cnxk_bphy, cnxk_bphy_cgx, intel_ioat_bdw, intel_ioat_skx, -intel_ioat_icx, intel_ntb_skx, intel_ntb_icx, -octeontx2_dma] +intel_ntb_skx, intel_ntb_icx, octeontx2_dma] I think the ioat_bdw and ioat_skx elements should also go down as DMA devices. With that change: Reviewed-by: Bruce Richardson I will change this in v5 and deprecate the rawdev IOAT driver. Thanks for the review Bruce, Conor.
Re: [dpdk-dev] [PATCH v4 4/5] test/crypto: add IV gen tests
> Subject: [PATCH v4 4/5] test/crypto: add IV gen tests > > From: Tejasree Kondoj > > Add test cases to verify IV generated by PMD. Title and description of the patch do not specify that This is for lookaside IPSec. Title: test/crypto: add IV gen cases for IPsec Description: Added cases to verify the IV generated by PMD for lookaside IPsec use case. It can also be mentioned that encap and decap are done One after the other so that known test vector is not required. Apart from that, Acked-by: Akhil Goyal > > Signed-off-by: Anoob Joseph > Signed-off-by: Tejasree Kondoj > ---
Re: [dpdk-dev] [PATCH v4 3/5] test/crypto: add lookaside IPsec ICV corrupt test case
> Subject: [PATCH v4 3/5] test/crypto: add lookaside IPsec ICV corrupt test case > > From: Tejasree Kondoj > > Adding lookaside IPsec ICV corrupt test case. Please elaborate the test case a bit more. There is no documentation for the test cases, we can Atleast add appropriate description. Apart from that, Acked-by: Akhil Goyal > > Signed-off-by: Anoob Joseph > Signed-off-by: Tejasree Kondoj >
Re: [dpdk-dev] [PATCH v4 01/11] dma/ioat: add device probe and removal functions
On Fri, Sep 17, 2021 at 03:42:17PM +, Conor Walsh wrote: Add the basic device probe/remove skeleton code and initial documentation for new IOAT DMA driver. Maintainers update is also included in this patch. Signed-off-by: Conor Walsh Reviewed-by: Kevin Laatz --- MAINTAINERS| 6 +++ doc/guides/dmadevs/index.rst | 2 + doc/guides/dmadevs/ioat.rst| 64 doc/guides/rel_notes/release_21_11.rst | 7 +-- drivers/dma/ioat/ioat_dmadev.c | 69 ++ drivers/dma/ioat/ioat_hw_defs.h| 35 + drivers/dma/ioat/ioat_internal.h | 20 drivers/dma/ioat/meson.build | 7 +++ drivers/dma/ioat/version.map | 3 ++ drivers/dma/meson.build| 1 + 10 files changed, 211 insertions(+), 3 deletions(-) create mode 100644 doc/guides/dmadevs/ioat.rst create mode 100644 drivers/dma/ioat/ioat_dmadev.c create mode 100644 drivers/dma/ioat/ioat_hw_defs.h create mode 100644 drivers/dma/ioat/ioat_internal.h create mode 100644 drivers/dma/ioat/meson.build create mode 100644 drivers/dma/ioat/version.map diff --git a/MAINTAINERS b/MAINTAINERS index 9cb59b831d..70993d23e8 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1209,6 +1209,12 @@ M: Kevin Laatz F: drivers/dma/idxd/ F: doc/guides/dmadevs/idxd.rst +Intel IOAT - EXPERIMENTAL +M: Bruce Richardson +M: Conor Walsh +F: drivers/dma/ioat/ +F: doc/guides/dmadevs/ioat.rst + Unlike the raw/ioat driver, this dmadev driver does not have a private APIs so I'm not sure it needs the EXPERIMENTAL tag on it. I will update this in v5. diff --git a/doc/guides/rel_notes/release_21_11.rst b/doc/guides/rel_notes/release_21_11.rst index c0bfd9c1ba..4d2b7bde1b 100644 --- a/doc/guides/rel_notes/release_21_11.rst +++ b/doc/guides/rel_notes/release_21_11.rst @@ -92,10 +92,11 @@ New Features * Device allocation and it's multi-process support. * Control and data plane functions. -* **Added IDXD dmadev driver implementation.** +* **Added Intel dmadev driver implementations.** - The IDXD dmadev driver provide device drivers for the Intel DSA devices. - This device driver can be used through the generic dmadev API. + The IDXD and IOAT dmadev drivers provide device drivers for Intel DSA + and IOAT devices. These device drivers can be used through the generic + dmadev API. I'm not sure about merging two driver additions into a single release note entry - one for the doc maintainers and tree committers to comment on. Rather than "IOAT devices" I think the official name of the hardware should be used, and you probably should add "respectively" at the end of the first sentence to make it clear (once you change the name) which hardware is used by which driver. I will separate them into 2 entries and reword in v5. Thanks, Conor.
Re: [dpdk-dev] [PATCH v2 3/5] eventdev/rx_adapter:add per queue event buffer configure support
On Tue, Sep 21, 2021 at 2:52 PM Naga Harish K S V
wrote:
>
> To configure per queue event buffer size, applications sets
> ``rte_event_eth_rx_adapter_params::use_queue_event_buf`` flag
> as true and is passed to ``rte_event_eth_rx_adapter_create_with_params``
> api.
>
> The per queue event buffer size is populated in
> ``rte_event_eth_rx_adapter_queue_conf::event_buf_size`` and passed
> to ``rte_event_eth_rx_adapter_queue_add`` api.
>
> Signed-off-by: Naga Harish K S V
Fix check-git-log issues
Wrong headline format:
eventdev/rx_adapter:add per queue event buffer configure support
Wrong headline case:
"test/event: add unit test for event buffer
size config api": api --> API
Wrong headline case:
"test/eventdev: add per rx queue event buffer
unit": rx --> Rx
Headline too long:
eventdev/rx_adapter: add support to configure event buffer size
eventdev/rx_adapter:add per queue event buffer configure support
Invalid patch(es) found - checked 5 patches
> ---
> .../prog_guide/event_ethernet_rx_adapter.rst | 19 ---
> lib/eventdev/rte_event_eth_rx_adapter.h | 4
> 2 files changed, 16 insertions(+), 7 deletions(-)
>
> diff --git a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> index dd753613bd..333e6f8192 100644
> --- a/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> +++ b/doc/guides/prog_guide/event_ethernet_rx_adapter.rst
> @@ -62,12 +62,14 @@ service function and needs to create an event port for
> it. The callback is
> expected to fill the ``struct rte_event_eth_rx_adapter_conf structure``
> passed to it.
>
> -If the application desires to control the event buffer size, it can use the
> -``rte_event_eth_rx_adapter_create_with_params()`` api. The event buffer size
> is
> -specified using ``struct rte_event_eth_rx_adapter_params::event_buf_size``.
> -The function is passed the event device to be associated with the adapter
> -and port configuration for the adapter to setup an event port if the
> -adapter needs to use a service function.
> +If the application desires to control the event buffer size at adapter level,
> +it can use the ``rte_event_eth_rx_adapter_create_with_params()`` api. The
> event
> +buffer size is specified using ``struct rte_event_eth_rx_adapter_params::
> +event_buf_size``. To configure the event buffer size at queue level, the
> boolean
> +flag ``struct rte_event_eth_rx_adapter_params::use_queue_event_buf`` need to
> be
> +set to true. The function is passed the event device to be associated with
> +the adapter and port configuration for the adapter to setup an event port
> +if the adapter needs to use a service function.
>
> Adding Rx Queues to the Adapter Instance
>
> @@ -79,7 +81,9 @@ parameter. Event information for packets from this Rx queue
> is encoded in the
> ``ev`` field of ``struct rte_event_eth_rx_adapter_queue_conf``. The
> servicing_weight member of the struct rte_event_eth_rx_adapter_queue_conf
> is the relative polling frequency of the Rx queue and is applicable when the
> -adapter uses a service core function.
> +adapter uses a service core function. The applications can configure queue
> +event buffer size in ``struct
> rte_event_eth_rx_adapter_queue_conf::event_buf_size``
> +parameter.
>
> .. code-block:: c
>
> @@ -90,6 +94,7 @@ adapter uses a service core function.
> queue_config.rx_queue_flags = 0;
> queue_config.ev = ev;
> queue_config.servicing_weight = 1;
> + queue_config.event_buf_size = 1024;
>
> err = rte_event_eth_rx_adapter_queue_add(id,
> eth_dev_id,
> diff --git a/lib/eventdev/rte_event_eth_rx_adapter.h
> b/lib/eventdev/rte_event_eth_rx_adapter.h
> index a7881097b4..b9f0563244 100644
> --- a/lib/eventdev/rte_event_eth_rx_adapter.h
> +++ b/lib/eventdev/rte_event_eth_rx_adapter.h
> @@ -199,6 +199,8 @@ struct rte_event_eth_rx_adapter_queue_conf {
> * Valid when RTE_EVENT_ETH_RX_ADAPTER_QUEUE_EVENT_VECTOR flag is set
> in
> * @see rte_event_eth_rx_adapter_queue_conf::rx_queue_flags.
> */
> + uint16_t event_buf_size;
> + /**< event buffer size for this queue */
> };
>
> /**
> @@ -265,6 +267,8 @@ struct rte_event_eth_rx_adapter_params {
> /**< size of event buffer for the adapter.
> * the size is aligned to BATCH_SIZE and added (2 * BATCH_SIZE)
> */
> + bool use_queue_event_buf;
> + /**< flag to indicate that event buffer is separate for each queue */
> };
>
> /**
> --
> 2.25.1
>
Re: [dpdk-dev] [PATCH v5 1/2] ethdev: make queue release callback optional
On 9/18/2021 1:35 PM, Xueming Li wrote: > Some drivers don't need Rx and Tx queue release callback, make them > optional. Clean up empty queue release callbacks for some drivers. > > Signed-off-by: Xueming Li > Reviewed-by: Andrew Rybchenko Acked-by: Ferruh Yigit
