RE: [PATCH 1/3] eal: fix pointer arithmetic with an expression argument
> From: Dmitry Kozlyuk [mailto:dmitry.kozl...@gmail.com]
> Sent: Sunday, 21 August 2022 22.50
>
> RTE_PTR_SUB(ptr, x) and RTE_PTR_ALIGN_FLOOR() worked incorrectly
> if "ptr" was an expression:
>
> uint32_t arr[3];
>
> RTE_PTR_SUB(arr + 1, sizeof(arr[0]));
> // expected: (uint32_t *)((uintptr_t)(arr + 1) - 4) == arr
> // actual: (uint32_t *)((uintptr_t) arr + 1 - 4) != arr
>
> RTE_PTR_ALIGN_FLOOR(arr + 2, sizeof(arr[0]));
> // expected: RTE_ALIGN_FLOOR((uintptr_t)(arr + 2), 4) == &arr[2]
> // actual: RTE_ALIGN_FLOOR((uintptr_t) arr + 2, 4) == &arr[0]
>
> Fix the macros and extend the relevant unit test.
Good catch. Serious bugs!
>
> Fixes: af75078fece3 ("first public release")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Dmitry Kozlyuk
> ---
> app/test/test_common.c | 11 +++
> lib/eal/include/rte_common.h | 4 ++--
> 2 files changed, 13 insertions(+), 2 deletions(-)
>
> diff --git a/app/test/test_common.c b/app/test/test_common.c
> index ef177cecb1..4194c1208a 100644
> --- a/app/test/test_common.c
> +++ b/app/test/test_common.c
> @@ -31,6 +31,7 @@ test_macros(int __rte_unused unused_parm)
>
> uintptr_t unused = 0;
> unsigned int smaller = SMALLER, bigger = BIGGER;
> + uint32_t arr[3];
>
> RTE_SET_USED(unused);
>
> @@ -41,6 +42,16 @@ test_macros(int __rte_unused unused_parm)
> FAIL_MACRO(RTE_PTR_ADD);
> if ((uintptr_t)RTE_PTR_SUB(BIGGER, PTR_DIFF) != SMALLER)
> FAIL_MACRO(RTE_PTR_SUB);
> + if (RTE_PTR_ADD(arr + 1, sizeof(arr[0])) != &arr[2])
> + FAIL_MACRO(RTE_PTR_ADD);
> + if (RTE_PTR_SUB(arr + 1, sizeof(arr[0])) != &arr[0])
> + FAIL_MACRO(RTE_PTR_SUB);
Very elegant test cases. :-)
> + if (RTE_PTR_ALIGN_FLOOR(arr + 2, 4) != &arr[2])
> + FAIL_MACRO(RTE_PTR_ALIGN_FLOOR);
> + if (RTE_PTR_ALIGN_CEIL(arr + 2, 4) != &arr[2])
> + FAIL_MACRO(RTE_PTR_ALIGN_CEIL);
While you are at it, consider adding a few more test cases, e.g.
RTE_PTR_ALIGN_FLOOR/CEIL(RTE_PTR_ADD(&arr[1], 1), 4), and
RTE_PTR_ALIGN_FLOOR/CEIL(RTE_PTR_ADD(&arr[1], sizeof(uint32_t) - 1), 4)
> + if (RTE_PTR_ALIGN(arr + 2, 4) != &arr[2])
> + FAIL_MACRO(RTE_PTR_ALIGN);
> if (RTE_PTR_DIFF(BIGGER, SMALLER) != PTR_DIFF)
> FAIL_MACRO(RTE_PTR_DIFF);
> if (RTE_MAX(SMALLER, BIGGER) != BIGGER)
> diff --git a/lib/eal/include/rte_common.h
> b/lib/eal/include/rte_common.h
> index a96cc2a138..d517e9f75f 100644
> --- a/lib/eal/include/rte_common.h
> +++ b/lib/eal/include/rte_common.h
> @@ -295,7 +295,7 @@ static void
> __attribute__((destructor(RTE_PRIO(prio)), used)) func(void)
> /**
> * subtract a byte-value offset from a pointer
> */
> -#define RTE_PTR_SUB(ptr, x) ((void*)((uintptr_t)ptr - (x)))
> +#define RTE_PTR_SUB(ptr, x) ((void *)((uintptr_t)(ptr) - (x)))
>
> /**
> * get the difference between two pointer values, i.e. how far apart
> @@ -320,7 +320,7 @@ static void
> __attribute__((destructor(RTE_PRIO(prio)), used)) func(void)
> * must be a power-of-two value.
> */
> #define RTE_PTR_ALIGN_FLOOR(ptr, align) \
> - ((typeof(ptr))RTE_ALIGN_FLOOR((uintptr_t)ptr, align))
> + ((typeof(ptr))RTE_ALIGN_FLOOR((uintptr_t)(ptr), align))
>
> /**
> * Macro to align a value to a given power-of-two. The resultant value
> --
> 2.33.1
>
Reviewed-by: Morten Brørup
RE: [PATCH 2/3] eal: uninline rte_str_to_size
> From: Dmitry Kozlyuk [mailto:dmitry.kozl...@gmail.com] > Sent: Sunday, 21 August 2022 22.50 > To: dev@dpdk.org > Cc: Dmitry Kozlyuk; Ray Kinsella > Subject: [PATCH 2/3] eal: uninline rte_str_to_size > > There is no reason for rte_str_to_size() to be inline. > Move the implementation out of . > Export it as a stable ABI because it always has been public. > > Signed-off-by: Dmitry Kozlyuk Acked-by: Morten Brørup > --- > Now doesn't need to #include and , > but removing them breaks some DPDK code, may break user code too. > I'm not sure what is the compatibility policy in this regard. > If such a breakage is allowed, I'd remove includes and fix DPDK code. > The question I'm asking myself here is: Do we want rte_common.h to include common headers like these, just so we don't need to include them elsewhere? I think not. I'm in favor of the principle of keeping it clean: Remove them from rte_common.h, and deal with the consequences. If we keep them, we will forget why they are there, and some day in the future, someone will ask what these unused headers are doing in .
RE: [PATCH 3/3] eal: deduplicate roundup code
> From: Dmitry Kozlyuk [mailto:dmitry.kozl...@gmail.com] > Sent: Sunday, 21 August 2022 22.50 > > RTE_CACHE_LINE_ROUNDUP() implementation repeated RTE_ALIGN_MUL_CEIL(). > In other places RTE_CACHE_LINE_SIZE is assumed to be a power-of-2, > so define RTE_CACHE_LINE_ROUNDUP() using RTE_ALIGN_CEIL(). > > Signed-off-by: Dmitry Kozlyuk > --- > lib/eal/include/rte_common.h | 4 +--- > 1 file changed, 1 insertion(+), 3 deletions(-) > > diff --git a/lib/eal/include/rte_common.h > b/lib/eal/include/rte_common.h > index 772e40f8c2..86c50c55e0 100644 > --- a/lib/eal/include/rte_common.h > +++ b/lib/eal/include/rte_common.h > @@ -425,9 +425,7 @@ rte_is_aligned(void *ptr, unsigned align) > #define RTE_CACHE_LINE_MASK (RTE_CACHE_LINE_SIZE-1) > > /** Return the first cache-aligned value greater or equal to size. */ > -#define RTE_CACHE_LINE_ROUNDUP(size) \ > - (RTE_CACHE_LINE_SIZE * ((size + RTE_CACHE_LINE_SIZE - 1) / \ > - RTE_CACHE_LINE_SIZE)) > +#define RTE_CACHE_LINE_ROUNDUP(size) RTE_ALIGN_CEIL(size, > RTE_CACHE_LINE_SIZE) > > /** Cache line size in terms of log2 */ > #if RTE_CACHE_LINE_SIZE == 64 > -- > 2.33.1 > Reviewed-by: Morten Brørup
[PATCH v1] vhost: fix build
This patch fixes the following build failure seen on CentOS 8
with gcc 12.1 because of uninitialized struct variable:
[..]
../lib/vhost/virtio_net.c:1159:18: warning: 'buf_vec[0].buf_addr' may be used
uninitialized [-Wmaybe-uninitialized]
1159 | buf_addr = buf_vec[vec_idx].buf_addr;
| ~^~~
[..]
Fixes: 873e8dad6f49 ("vhost: support packed ring in async datapath")
Signed-off-by: Min Zhou
---
lib/vhost/virtio_net.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c
index 35fa4670fd..4878bb2d8a 100644
--- a/lib/vhost/virtio_net.c
+++ b/lib/vhost/virtio_net.c
@@ -1837,6 +1837,7 @@ virtio_dev_rx_async_packed(struct virtio_net *dev, struct
vhost_virtqueue *vq,
{
struct buf_vector buf_vec[BUF_VECTOR_MAX];
+ memset(buf_vec, 0, sizeof(buf_vec));
if (unlikely(vhost_enqueue_async_packed(dev, vq, pkt, buf_vec,
nr_descs, nr_buffers) < 0)) {
VHOST_LOG_DATA(dev->ifname, DEBUG, "failed to get enough desc
from vring\n");
--
2.31.1
RE: [PATCH 0/2] IPsec on Arm
Hi Ruifeng,
We have no technical issues with these patches based on their current
functionality. However, it is worth pointing out that we are planning some
changes to the Intel® Multi-Buffer Crypto for IPSec library which will require
API changes in the future. These changes are, but not limited to, to
consolidate some of the crypto PMDs into a single PMD, which will simplify the
code and reduce future maintenance effort. When these changes are made, your
library will need to be updated too in order to take advantage of the
consolidation. We can provide further details of the changes that we’re
considering separately, so that you’re not taken by surprise when these change
occurs.
We also want you to know, for any change we will make to these PMDs, we do not
guarantee it works on ARM platform. We advise you to do necessary
testing/verification in necessary testing/verification to the future patches
for the PMDs based on ipsec-mb on your platform.
Regards,
Fan
> -Original Message-
> From: Ruifeng Wang
> Sent: Wednesday, July 27, 2022 9:14 AM
> To: Zhang, Roy Fan ; De Lara Guarch, Pablo
> ; Wang, Yipeng1
> ; Gobriel, Sameh ;
> Richardson, Bruce ; Medvedkin, Vladimir
> ; gak...@marvell.com
> Cc: dev@dpdk.org; honnappa.nagaraha...@arm.com; n...@arm.com; Ruifeng
> Wang
> Subject: [PATCH 0/2] IPsec on Arm
>
> This patch set integrates Arm port of ipsec_mb library.
> ZUC and SNOW3g PMDs are available on Arm platform with this integration.
>
> This series depends on:
> http://patches.dpdk.org/project/dpdk/patch/20220630154123.2565439-1-
> asek...@marvell.com/
>
> Depends-on: patch-113578 ("crypto/ipsec_mb: enable support for arm64")
>
> Ruifeng Wang (2):
> crypto/ipsec_mb: remove redundant includes
> crypto/ipsec_mb: enable IPsec on Arm platform
>
> app/test/test_cryptodev_hash_test_vectors.h | 4
> doc/guides/cryptodevs/snow3g.rst | 14 ++
> doc/guides/cryptodevs/zuc.rst| 14 ++
> drivers/common/qat/meson.build | 6 +-
> drivers/crypto/ipsec_mb/ipsec_mb_private.c | 6 ++
> drivers/crypto/ipsec_mb/ipsec_mb_private.h | 4
> drivers/crypto/ipsec_mb/meson.build | 6 +-
> drivers/crypto/ipsec_mb/pmd_aesni_gcm_priv.h | 2 --
> drivers/crypto/ipsec_mb/pmd_aesni_mb_priv.h | 2 --
> drivers/crypto/qat/qat_sym_session.c | 4
> 10 files changed, 48 insertions(+), 14 deletions(-)
>
> --
> 2.25.1
Re: [dpdk-kmods v2] windows/netuio: fix BAR parsing
在 2022/8/12 6:17, Pallavi Kadam 写道: Current code was always checking the 'prev_bar & PCI_TYPE_64BIT' though only the first BAR slot of a 64-bit BAR contains flags. Also for certain PCIe devices, BAR values were not continuous. This patch fixes this incorrectness and maps the BAR addresses correctly. Reported-by: Qiao Liu Suggested-by: Dmitry Kozlyuk Signed-off-by: Dmitry Kozlyuk Tested-by: Pallavi Kadam --- Acked-by: Qiao Liu
[PATCH] sched:subport field is unused in hqos profile.
--- examples/qos_sched/profile.cfg | 2 -- 1 file changed, 2 deletions(-) diff --git a/examples/qos_sched/profile.cfg b/examples/qos_sched/profile.cfg index d4b21c0170..8da5777538 100644 --- a/examples/qos_sched/profile.cfg +++ b/examples/qos_sched/profile.cfg @@ -26,8 +26,6 @@ number of subports per port = 1 number of pipes per subport = 4096 queue sizes = 64 64 64 64 64 64 64 64 64 64 64 64 64 -subport 0-8 = 0; These subports are configured with subport profile 0 - [subport profile 0] tb rate = 125000 ; Bytes per second tb size = 100 ; Bytes -- 2.25.1
[PATCH] sched:In rte_sched_subport_config() API, subport_profile_id is not set correctly.
--- lib/sched/rte_sched.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/lib/sched/rte_sched.c b/lib/sched/rte_sched.c index 599c7e9536..09f855a04b 100644 --- a/lib/sched/rte_sched.c +++ b/lib/sched/rte_sched.c @@ -1263,8 +1263,6 @@ rte_sched_subport_config(struct rte_sched_port *port, n_subports++; - subport_profile_id = 0; - /* Port */ port->subports[subport_id] = s; -- 2.25.1
[PATCH] sched:higher rates of pipe and subport are not taken care in application.
---
examples/qos_sched/cfg_file.c | 66 +--
1 file changed, 33 insertions(+), 33 deletions(-)
diff --git a/examples/qos_sched/cfg_file.c b/examples/qos_sched/cfg_file.c
index 450482f07d..fe0b42b023 100644
--- a/examples/qos_sched/cfg_file.c
+++ b/examples/qos_sched/cfg_file.c
@@ -62,71 +62,71 @@ cfg_load_pipe(struct rte_cfgfile *cfg, struct
rte_sched_pipe_params *pipe_params
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tb rate");
if (entry)
- pipe_params[j].tb_rate = (uint64_t)atoi(entry);
+ pipe_params[j].tb_rate = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tb size");
if (entry)
- pipe_params[j].tb_size = (uint64_t)atoi(entry);
+ pipe_params[j].tb_size = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc period");
if (entry)
- pipe_params[j].tc_period = (uint64_t)atoi(entry);
+ pipe_params[j].tc_period = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 0 rate");
if (entry)
- pipe_params[j].tc_rate[0] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[0] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 1 rate");
if (entry)
- pipe_params[j].tc_rate[1] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[1] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 2 rate");
if (entry)
- pipe_params[j].tc_rate[2] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[2] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 3 rate");
if (entry)
- pipe_params[j].tc_rate[3] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[3] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 4 rate");
if (entry)
- pipe_params[j].tc_rate[4] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[4] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 5 rate");
if (entry)
- pipe_params[j].tc_rate[5] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[5] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 6 rate");
if (entry)
- pipe_params[j].tc_rate[6] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[6] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 7 rate");
if (entry)
- pipe_params[j].tc_rate[7] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[7] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 8 rate");
if (entry)
- pipe_params[j].tc_rate[8] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[8] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 9 rate");
if (entry)
- pipe_params[j].tc_rate[9] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[9] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 10 rate");
if (entry)
- pipe_params[j].tc_rate[10] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[10] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 11 rate");
if (entry)
- pipe_params[j].tc_rate[11] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[11] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 12 rate");
if (entry)
- pipe_params[j].tc_rate[12] = (uint64_t)atoi(entry);
+ pipe_params[j].tc_rate[12] = (uint64_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 12
oversubscription weight");
if (entry)
- pipe_params[j].tc_ov_weight = (uint8_t)atoi(entry);
+ pipe_params[j].tc_ov_weight = (uint8_t)atol(entry);
entry = rte_cfgfile_get_entry(cfg, pipe_name, "tc 12 wrr
weights");
if (entry) {
@@ -163,67 +163,67 @@ cfg_load_subport_profile(struct rte_cfgfile *cfg,
entry = rte_cfgfile_get_entry(c
[PATCH] linux/igb_uio: make module parameters visible in sysfs
- explicitly set default value for intr_mode
- give read permission on wc_activate
- applying patch gives below visibility
```
$ more /sys/module/igb_uio/parameters/* | cat
::
/sys/module/igb_uio/parameters/intr_mode
::
msix
::
/sys/module/igb_uio/parameters/wc_activate
::
0
```
Signed-off-by: Shinae Woo
---
linux/igb_uio/igb_uio.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/linux/igb_uio/igb_uio.c b/linux/igb_uio/igb_uio.c
index 33e0e02..c927ae6 100644
--- a/linux/igb_uio/igb_uio.c
+++ b/linux/igb_uio/igb_uio.c
@@ -44,7 +44,7 @@ struct rte_uio_pci_dev {
};
static int wc_activate;
-static char *intr_mode;
+static char *intr_mode = RTE_INTR_MODE_MSIX_NAME;
static enum rte_intr_mode igbuio_intr_mode_preferred = RTE_INTR_MODE_MSIX;
/* sriov sysfs */
static ssize_t
@@ -663,7 +663,7 @@ MODULE_PARM_DESC(intr_mode,
"" RTE_INTR_MODE_LEGACY_NAME " Use Legacy interrupt\n"
"\n");
-module_param(wc_activate, int, 0);
+module_param(wc_activate, int, S_IRUGO);
MODULE_PARM_DESC(wc_activate,
"Activate support for write combining (WC) (default=0)\n"
"0 - disable\n"
--
2.20.1
RE: [PATCH 0/2] IPsec on Arm
> -Original Message-
> From: Zhang, Roy Fan
> Sent: Monday, August 22, 2022 3:54 PM
> To: Ruifeng Wang ; De Lara Guarch, Pablo
> ; Wang, Yipeng1 ;
> Gobriel, Sameh
> ; Richardson, Bruce ;
> Medvedkin,
> Vladimir ; gak...@marvell.com
> Cc: dev@dpdk.org; Honnappa Nagarahalli ; nd
>
> Subject: RE: [PATCH 0/2] IPsec on Arm
>
> Hi Ruifeng,
Hi Fan,
>
> We have no technical issues with these patches based on their current
> functionality.
> However, it is worth pointing out that we are planning some changes to the
> Intel(r) Multi-
> Buffer Crypto for IPSec library which will require API changes in the future.
> These
> changes are, but not limited to, to consolidate some of the crypto PMDs into
> a single PMD,
> which will simplify the code and reduce future maintenance effort. When these
> changes are
> made, your library will need to be updated too in order to take advantage of
> the
Thanks for the heads-up. Really appreciate it.
> consolidation. We can provide further details of the changes that we're
> considering
> separately, so that you're not taken by surprise when these change occurs.
Yep, let's communicate this in a separate thread.
>
> We also want you to know, for any change we will make to these PMDs, we do
> not guarantee
> it works on ARM platform. We advise you to do necessary testing/verification
> in necessary
> testing/verification to the future patches for the PMDs based on ipsec-mb on
> your platform.
Yes, I will verify ZUC/SNOW3g PMD patches on Arm platform.
>
> Regards,
> Fan
>
> > -Original Message-
> > From: Ruifeng Wang
> > Sent: Wednesday, July 27, 2022 9:14 AM
> > To: Zhang, Roy Fan ; De Lara Guarch, Pablo
> > ; Wang, Yipeng1
> > ; Gobriel, Sameh ;
> > Richardson, Bruce ; Medvedkin, Vladimir
> > ; gak...@marvell.com
> > Cc: dev@dpdk.org; honnappa.nagaraha...@arm.com; n...@arm.com; Ruifeng
> > Wang
> > Subject: [PATCH 0/2] IPsec on Arm
> >
> > This patch set integrates Arm port of ipsec_mb library.
> > ZUC and SNOW3g PMDs are available on Arm platform with this integration.
> >
> > This series depends on:
> > http://patches.dpdk.org/project/dpdk/patch/20220630154123.2565439-1-
> > asek...@marvell.com/
> >
> > Depends-on: patch-113578 ("crypto/ipsec_mb: enable support for arm64")
> >
> > Ruifeng Wang (2):
> > crypto/ipsec_mb: remove redundant includes
> > crypto/ipsec_mb: enable IPsec on Arm platform
> >
> > app/test/test_cryptodev_hash_test_vectors.h | 4
> > doc/guides/cryptodevs/snow3g.rst | 14 ++
> > doc/guides/cryptodevs/zuc.rst| 14 ++
> > drivers/common/qat/meson.build | 6 +-
> > drivers/crypto/ipsec_mb/ipsec_mb_private.c | 6 ++
> > drivers/crypto/ipsec_mb/ipsec_mb_private.h | 4
> > drivers/crypto/ipsec_mb/meson.build | 6 +-
> > drivers/crypto/ipsec_mb/pmd_aesni_gcm_priv.h | 2 --
> > drivers/crypto/ipsec_mb/pmd_aesni_mb_priv.h | 2 --
> > drivers/crypto/qat/qat_sym_session.c | 4
> > 10 files changed, 48 insertions(+), 14 deletions(-)
> >
> > --
> > 2.25.1
[PATCH v3 0/3] net/octeon_ep: rename driver and add features
This patch set renames the net/octeontx_ep driver to net/octeon_ep
and 2nd and 3rd patches add support for basic stats and link status.
Changes in v3:
- Updated commit messaage.
- Updated deprecation.rst.
Changes in v2:
Added new features in the renamed driver.
Sathesh Edara (3):
net/octeontx_ep: rename as octeon_ep
net/octeon_ep: support basic stats
net/octeon_ep: support link status
MAINTAINERS | 6 +-
.../{octeontx_ep.ini => octeon_ep.ini}| 4 +-
doc/guides/nics/index.rst | 2 +-
.../nics/{octeontx_ep.rst => octeon_ep.rst} | 4 +-
doc/guides/rel_notes/deprecation.rst | 5 --
drivers/net/meson.build | 2 +-
.../{octeontx_ep => octeon_ep}/meson.build| 0
.../{octeontx_ep => octeon_ep}/otx2_ep_vf.c | 0
.../{octeontx_ep => octeon_ep}/otx2_ep_vf.h | 0
.../otx_ep_common.h | 0
.../otx_ep_ethdev.c | 69 +++
.../{octeontx_ep => octeon_ep}/otx_ep_rxtx.c | 0
.../{octeontx_ep => octeon_ep}/otx_ep_rxtx.h | 0
.../{octeontx_ep => octeon_ep}/otx_ep_vf.c| 0
.../{octeontx_ep => octeon_ep}/otx_ep_vf.h| 0
.../{octeontx_ep => octeon_ep}/version.map| 0
16 files changed, 79 insertions(+), 13 deletions(-)
rename doc/guides/nics/features/{octeontx_ep.ini => octeon_ep.ini} (64%)
rename doc/guides/nics/{octeontx_ep.rst => octeon_ep.rst} (87%)
rename drivers/net/{octeontx_ep => octeon_ep}/meson.build (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_common.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_ethdev.c (86%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/version.map (100%)
--
2.36.1
[PATCH v3 2/3] net/octeon_ep: support basic stats
Added functionality to fetch and reset ethdev stats.
Signed-off-by: Sathesh Edara
---
doc/guides/nics/features/octeon_ep.ini | 1 +
drivers/net/octeon_ep/otx_ep_ethdev.c | 52 ++
2 files changed, 53 insertions(+)
diff --git a/doc/guides/nics/features/octeon_ep.ini
b/doc/guides/nics/features/octeon_ep.ini
index 141d918466..b304ff8877 100644
--- a/doc/guides/nics/features/octeon_ep.ini
+++ b/doc/guides/nics/features/octeon_ep.ini
@@ -8,4 +8,5 @@ Speed capabilities = P
SR-IOV = Y
Linux= Y
x86-64 = Y
+Basic stats = Y
Usage doc= Y
diff --git a/drivers/net/octeon_ep/otx_ep_ethdev.c
b/drivers/net/octeon_ep/otx_ep_ethdev.c
index 806add246b..cb45bd7a8a 100644
--- a/drivers/net/octeon_ep/otx_ep_ethdev.c
+++ b/drivers/net/octeon_ep/otx_ep_ethdev.c
@@ -337,6 +337,56 @@ otx_ep_tx_queue_release(struct rte_eth_dev *dev, uint16_t
q_no)
otx_ep_delete_iqs(tq->otx_ep_dev, tq->q_no);
}
+static int
+otx_ep_dev_stats_reset(struct rte_eth_dev *dev)
+{
+ struct otx_ep_device *otx_epvf = OTX_EP_DEV(dev);
+ uint32_t i;
+
+ for (i = 0; i < otx_epvf->nb_tx_queues; i++)
+ memset(&otx_epvf->instr_queue[i]->stats, 0,
+ sizeof(struct otx_ep_iq_stats));
+
+ for (i = 0; i < otx_epvf->nb_rx_queues; i++)
+ memset(&otx_epvf->droq[i]->stats, 0,
+ sizeof(struct otx_ep_droq_stats));
+
+ return 0;
+}
+
+static int
+otx_ep_dev_stats_get(struct rte_eth_dev *eth_dev,
+ struct rte_eth_stats *stats)
+{
+ struct otx_ep_device *otx_epvf = OTX_EP_DEV(eth_dev);
+ struct otx_ep_iq_stats *ostats;
+ struct otx_ep_droq_stats *istats;
+ uint32_t i;
+
+ memset(stats, 0, sizeof(struct rte_eth_stats));
+
+ for (i = 0; i < otx_epvf->nb_tx_queues; i++) {
+ ostats = &otx_epvf->instr_queue[i]->stats;
+ stats->q_opackets[i] = ostats->tx_pkts;
+ stats->q_obytes[i] = ostats->tx_bytes;
+ stats->opackets += ostats->tx_pkts;
+ stats->obytes += ostats->tx_bytes;
+ stats->oerrors += ostats->instr_dropped;
+ }
+ for (i = 0; i < otx_epvf->nb_rx_queues; i++) {
+ istats = &otx_epvf->droq[i]->stats;
+ stats->q_ipackets[i] = istats->pkts_received;
+ stats->q_ibytes[i] = istats->bytes_received;
+ stats->q_errors[i] = istats->rx_err;
+ stats->ipackets += istats->pkts_received;
+ stats->ibytes += istats->bytes_received;
+ stats->imissed += istats->rx_alloc_failure;
+ stats->ierrors += istats->rx_err;
+ stats->rx_nombuf += istats->rx_alloc_failure;
+ }
+ return 0;
+}
+
/* Define our ethernet definitions */
static const struct eth_dev_ops otx_ep_eth_dev_ops = {
.dev_configure = otx_ep_dev_configure,
@@ -347,6 +397,8 @@ static const struct eth_dev_ops otx_ep_eth_dev_ops = {
.tx_queue_setup = otx_ep_tx_queue_setup,
.tx_queue_release = otx_ep_tx_queue_release,
.dev_infos_get = otx_ep_dev_info_get,
+ .stats_get = otx_ep_dev_stats_get,
+ .stats_reset= otx_ep_dev_stats_reset,
};
static int
--
2.36.1
[PATCH v3 3/3] net/octeon_ep: support link status
Added functionality to update link speed, duplex mode and link state.
Signed-off-by: Sathesh Edara
---
doc/guides/nics/features/octeon_ep.ini | 1 +
drivers/net/octeon_ep/otx_ep_ethdev.c | 17 +
2 files changed, 18 insertions(+)
diff --git a/doc/guides/nics/features/octeon_ep.ini
b/doc/guides/nics/features/octeon_ep.ini
index b304ff8877..305e219262 100644
--- a/doc/guides/nics/features/octeon_ep.ini
+++ b/doc/guides/nics/features/octeon_ep.ini
@@ -9,4 +9,5 @@ SR-IOV = Y
Linux= Y
x86-64 = Y
Basic stats = Y
+Link status = Y
Usage doc= Y
diff --git a/drivers/net/octeon_ep/otx_ep_ethdev.c
b/drivers/net/octeon_ep/otx_ep_ethdev.c
index cb45bd7a8a..77def6daa1 100644
--- a/drivers/net/octeon_ep/otx_ep_ethdev.c
+++ b/drivers/net/octeon_ep/otx_ep_ethdev.c
@@ -387,6 +387,22 @@ otx_ep_dev_stats_get(struct rte_eth_dev *eth_dev,
return 0;
}
+static int
+otx_ep_dev_link_update(struct rte_eth_dev *eth_dev, int wait_to_complete)
+{
+ RTE_SET_USED(wait_to_complete);
+
+ if (!eth_dev->data->dev_started)
+ return 0;
+ struct rte_eth_link link;
+
+ memset(&link, 0, sizeof(link));
+ link.link_status = RTE_ETH_LINK_UP;
+ link.link_speed = RTE_ETH_SPEED_NUM_10G;
+ link.link_duplex = RTE_ETH_LINK_FULL_DUPLEX;
+ return rte_eth_linkstatus_set(eth_dev, &link);
+}
+
/* Define our ethernet definitions */
static const struct eth_dev_ops otx_ep_eth_dev_ops = {
.dev_configure = otx_ep_dev_configure,
@@ -399,6 +415,7 @@ static const struct eth_dev_ops otx_ep_eth_dev_ops = {
.dev_infos_get = otx_ep_dev_info_get,
.stats_get = otx_ep_dev_stats_get,
.stats_reset= otx_ep_dev_stats_reset,
+ .link_update= otx_ep_dev_link_update,
};
static int
--
2.36.1
[PATCH v3 1/3] net/octeontx_ep: rename as octeon_ep
This patch renames octeon end point driver from octeontx_ep to
octeon_ep to enable single unified driver to support current
OcteonTx and future Octeon PCI endpoint NICs to reflect common
driver for all Octeon based PCI endpoint NICs.
Signed-off-by: Sathesh Edara
---
MAINTAINERS | 6 +++---
doc/guides/nics/features/{octeontx_ep.ini => octeon_ep.ini} | 2 +-
doc/guides/nics/index.rst | 2 +-
doc/guides/nics/{octeontx_ep.rst => octeon_ep.rst} | 4 ++--
doc/guides/rel_notes/deprecation.rst| 5 -
drivers/net/meson.build | 2 +-
drivers/net/{octeontx_ep => octeon_ep}/meson.build | 0
drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.c | 0
drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.h | 0
drivers/net/{octeontx_ep => octeon_ep}/otx_ep_common.h | 0
drivers/net/{octeontx_ep => octeon_ep}/otx_ep_ethdev.c | 0
drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.c| 0
drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.h| 0
drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.c | 0
drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.h | 0
drivers/net/{octeontx_ep => octeon_ep}/version.map | 0
16 files changed, 8 insertions(+), 13 deletions(-)
rename doc/guides/nics/features/{octeontx_ep.ini => octeon_ep.ini} (75%)
rename doc/guides/nics/{octeontx_ep.rst => octeon_ep.rst} (87%)
rename drivers/net/{octeontx_ep => octeon_ep}/meson.build (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_common.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_ethdev.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.c (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.h (100%)
rename drivers/net/{octeontx_ep => octeon_ep}/version.map (100%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 32ffdd1a61..d7c7fa4cdf 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -812,9 +812,9 @@ M: Radha Mohan Chintakuntla
M: Veerasenareddy Burru
M: Sathesh Edara
T: git://dpdk.org/next/dpdk-next-net-mrvl
-F: drivers/net/octeontx_ep/
-F: doc/guides/nics/features/octeontx_ep.ini
-F: doc/guides/nics/octeontx_ep.rst
+F: drivers/net/octeon_ep/
+F: doc/guides/nics/features/octeon_ep.ini
+F: doc/guides/nics/octeon_ep.rst
Mellanox mlx4
M: Matan Azrad
diff --git a/doc/guides/nics/features/octeontx_ep.ini
b/doc/guides/nics/features/octeon_ep.ini
similarity index 75%
rename from doc/guides/nics/features/octeontx_ep.ini
rename to doc/guides/nics/features/octeon_ep.ini
index d1453f5bee..141d918466 100644
--- a/doc/guides/nics/features/octeontx_ep.ini
+++ b/doc/guides/nics/features/octeon_ep.ini
@@ -1,5 +1,5 @@
;
-; Supported features of the 'octeontx_ep' network poll mode driver.
+; Supported features of the 'octeon_ep' network poll mode driver.
;
; Refer to default.ini for the full list of available PMD features.
;
diff --git a/doc/guides/nics/index.rst b/doc/guides/nics/index.rst
index f48e9f815c..f80906a97d 100644
--- a/doc/guides/nics/index.rst
+++ b/doc/guides/nics/index.rst
@@ -52,7 +52,7 @@ Network Interface Controller Drivers
ngbe
null
octeontx
-octeontx_ep
+octeon_ep
pfe
qede
sfc_efx
diff --git a/doc/guides/nics/octeontx_ep.rst b/doc/guides/nics/octeon_ep.rst
similarity index 87%
rename from doc/guides/nics/octeontx_ep.rst
rename to doc/guides/nics/octeon_ep.rst
index 2ec8a034b5..b5040aeee2 100644
--- a/doc/guides/nics/octeontx_ep.rst
+++ b/doc/guides/nics/octeon_ep.rst
@@ -4,9 +4,9 @@
OCTEON TX EP Poll Mode driver
=
-The OCTEON TX EP ETHDEV PMD (**librte_pmd_octeontx_ep**) provides poll mode
+The OCTEON TX EP ETHDEV PMD (**librte_pmd_octeon_ep**) provides poll mode
ethdev driver support for the virtual functions (VF) of **Marvell OCTEON 9**
-and **Cavium OCTEON TX** families of adapters in SR-IOV context.
+and **Cavium OCTEON** families of adapters in SR-IOV context.
More information can be found at `Marvell Official Website
`_.
diff --git a/doc/guides/rel_notes/deprecation.rst
b/doc/guides/rel_notes/deprecation.rst
index e7583cae4c..6558bd7003 100644
--- a/doc/guides/rel_notes/deprecation.rst
+++ b/doc/guides/rel_notes/deprecation.rst
@@ -220,11 +220,6 @@ Deprecation Notices
Names for the telemetry commands will be similarly limited.
The parameters for telemetry commands are unaffected by this change.
-* net/octeontx_ep: The driver ``octeontx_ep`` was to support OCTEON TX
- line of produ
[Bug 1070] [dpdk-20.11.6-rc1] unit_tests_eal/link_bonding_rssconf: link_bonding_rssconf_autotest test failed
https://bugs.dpdk.org/show_bug.cgi?id=1070 Bug ID: 1070 Summary: [dpdk-20.11.6-rc1] unit_tests_eal/link_bonding_rssconf: link_bonding_rssconf_autotest test failed Product: DPDK Version: 20.11 Hardware: All OS: All Status: UNCONFIRMED Severity: normal Priority: Normal Component: testpmd Assignee: dev@dpdk.org Reporter: linglix.c...@intel.com Target Milestone: --- Environment OS: Ubuntu 18.04.3 LTS/4.15.0-55-generic Compiler: gcc version (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 Hardware platform: Intel(R) Xeon(R) Gold 6139 CPU @ 2.30GHz NIC hardware: Ethernet Controller XXV710 for 25GbE SFP28 158b NIC firmware: 8.00 0x80008c1a 1.2766.0 driver: vfio-pci kdriver: i40e-2.14.13 Test Setup 1. launch testpmd x86_64-native-linuxapp-gcc/app/test/dpdk-test -l 1,2,3,4 -n 4 2.start autotest RTE>>link_bonding_rssconf_autotest Show the output from the previous commands. + --- + + Test Suite : RSS Dynamic Configuration for Bonding Unit Test Suite + --- + + TestCase [ 0] : test_setup succeeded Device with port_id=0 already stopped Device with port_id=1 already stopped Device with port_id=2 already stopped Device with port_id=3 already stopped bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 0: Operation not supported bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 1: Operation not supported bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 2: Operation not supported bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 3: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 0: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 1: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 2: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 3: Operation not supported bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated bond_ethdev_rss_hash_update(3124) - rss_key will be truncated Port 0 must be stopped to allow reset Port 0 must be stopped to allow reset Port 1 must be stopped to allow reset Port 2 must be stopped to allow reset Port 3 must be stopped to allow reset + TestCase [ 1] : test_rss succeeded bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 0: Operation not supported bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 1: Operation not supported bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 2: Operation not supported bond_ethdev_promiscuous_disable(2669) - Failed to disable promiscuous mode for port 3: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 0: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 1: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 2: Operation not supported bond_ethdev_allmulticast_disable(2815) - Failed to disable allmulti mode for port 3: Operation not supported EAL: Test assert test_rss_config_lazy line 502 failed: Succeeded in setting bonded port hash function EAL: Test assert test_rss_lazy line 568 failed: Succeeded in setting RSS hash when RX_RSS mq_mode is turned off Port 0 must be stopped to allow reset Port 1 must be stopped to allow reset Port 2 must be stopped to allow reset Port 3 must be stopped to allow reset + TestCase [ 2] : test_rss_lazy failed Device with port_id=4 already stopped + --- + + Test Suite Summary + Tests Total :3 + Tests Skipped : 0 + Tests Executed : 3 + Tests Unsupported: 0 + Tests Passed : 2 + Tests Failed : 1 + --- + Test Failed Expected Result + --- + + Test Suite Summary : RSS Dynamic Configuration for Bonding Unit Test Suite + --- + + Tests Total :3 + Tests Skipped :
[PATCH v4] examples/vm_power_manager: use safe version of list iterator
From: Hamza Khan
Currently, when vm_power_manager exits, we are using a LIST_FOREACH
macro to iterate over VM info structures while freeing them. This
leads to use-after-free error. To address this, replace all usages of
LIST_* with TAILQ_* macros, and use the RTE_TAILQ_FOREACH_SAFE macro
to iterate and delete VM info structures.
* The change is small and doesn’t affect other code
* Testing was performed on the patch
Fixes: e8ae9b662506 ("examples/vm_power: channel manager and monitor in host")
Cc: alan.ca...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Hamza Khan
Reviewed-by: Reshma Pattan
Acked-by: Reshma Pattan
Signed-off-by: Reshma Pattan
---
v4: fix header file inclusion
---
examples/vm_power_manager/channel_manager.c | 20 +++-
1 file changed, 11 insertions(+), 9 deletions(-)
diff --git a/examples/vm_power_manager/channel_manager.c
b/examples/vm_power_manager/channel_manager.c
index 838465ab4b..cb872ad2d5 100644
--- a/examples/vm_power_manager/channel_manager.c
+++ b/examples/vm_power_manager/channel_manager.c
@@ -22,6 +22,7 @@
#include
#include
#include
+#include
#include
@@ -30,6 +31,7 @@
#include "power_manager.h"
+
#define RTE_LOGTYPE_CHANNEL_MANAGER RTE_LOGTYPE_USER1
struct libvirt_vm_info lvm_info[MAX_CLIENTS];
@@ -58,16 +60,16 @@ struct virtual_machine_info {
virDomainInfo info;
rte_spinlock_t config_spinlock;
int allow_query;
- LIST_ENTRY(virtual_machine_info) vms_info;
+ RTE_TAILQ_ENTRY(virtual_machine_info) vms_info;
};
-LIST_HEAD(, virtual_machine_info) vm_list_head;
+RTE_TAILQ_HEAD(, virtual_machine_info) vm_list_head;
static struct virtual_machine_info *
find_domain_by_name(const char *name)
{
struct virtual_machine_info *info;
- LIST_FOREACH(info, &vm_list_head, vms_info) {
+ RTE_TAILQ_FOREACH(info, &vm_list_head, vms_info) {
if (!strncmp(info->name, name, CHANNEL_MGR_MAX_NAME_LEN-1))
return info;
}
@@ -878,7 +880,7 @@ add_vm(const char *vm_name)
new_domain->allow_query = 0;
rte_spinlock_init(&(new_domain->config_spinlock));
- LIST_INSERT_HEAD(&vm_list_head, new_domain, vms_info);
+ TAILQ_INSERT_HEAD(&vm_list_head, new_domain, vms_info);
return 0;
}
@@ -900,7 +902,7 @@ remove_vm(const char *vm_name)
rte_spinlock_unlock(&vm_info->config_spinlock);
return -1;
}
- LIST_REMOVE(vm_info, vms_info);
+ TAILQ_REMOVE(&vm_list_head, vm_info, vms_info);
rte_spinlock_unlock(&vm_info->config_spinlock);
rte_free(vm_info);
return 0;
@@ -953,7 +955,7 @@ channel_manager_init(const char *path __rte_unused)
{
virNodeInfo info;
- LIST_INIT(&vm_list_head);
+ TAILQ_INIT(&vm_list_head);
if (connect_hypervisor(path) < 0) {
global_n_host_cpus = 64;
global_hypervisor_available = 0;
@@ -1005,9 +1007,9 @@ channel_manager_exit(void)
{
unsigned i;
char mask[RTE_MAX_LCORE];
- struct virtual_machine_info *vm_info;
+ struct virtual_machine_info *vm_info, *tmp;
- LIST_FOREACH(vm_info, &vm_list_head, vms_info) {
+ RTE_TAILQ_FOREACH_SAFE(vm_info, &vm_list_head, vms_info, tmp) {
rte_spinlock_lock(&(vm_info->config_spinlock));
@@ -1022,7 +1024,7 @@ channel_manager_exit(void)
}
rte_spinlock_unlock(&(vm_info->config_spinlock));
- LIST_REMOVE(vm_info, vms_info);
+ TAILQ_REMOVE(&vm_list_head, vm_info, vms_info);
rte_free(vm_info);
}
--
2.25.1
Re: Reason to alway to build both static and shared libs
On Wed, Aug 03, 2022 at 10:13:24PM -0700, Jianshen Liu wrote: >Hi all, Could I know the reason for always building both static and >shared libs of DPDK? I can find the [1]patch to enable this behavior, >but it seems that it didn't mention the reason behind it. Also, if I >propose a change to use "both" as the default for default_library in >meson's config file and still allow users to choose either static or >shared as they want, is there any reason against that change? Thanks, >Jianshen > When DPDK moved to use the meson build system, there was no "both" option for static and shared libs, so that is the primary reason why it is not used as the default. As for why even without that both libraries are always built, the main reasons we do so are: * traditionally DPDK has been built and linked into applications as a static library, and when moving build systems from make to meson we wanted to keep that as the default, since the transition was already confusing enough for users. * with the older make build system when users submitted patches to DPDK there was many, many cases where the patches did not include changes to the version.map files, leading to shared lib build failures. By ensuring that shared libs are always built, this whole class of patch errors are caught before users push their patches to the mailing list. Therefore by always building both shared and static we can keep consistency and ensure higher quality patches to the mailing list. I have looked a little into this in the past and from that investigation some other considerations are: * We can't actually use the built-in "both_libraries" option in DPDK directly since - due to ABI versioning - some files actually need to be compiled with different flags for static libraries (which don't have function versions) and shared libs (which do). This means that even if we do use "both_library" function, or "library" function with a default of "both" we will still need to maintain the code path fallback for building static and shared separately. * The built-in option for building static and/or shared libraries only controls what libraries get built - it does not control what way things get linked. Right now, the default in DPDK is to build both libraries but link the applications like testpmd or examples, using the static libs. If we use the defaults in meson for "both" the shared library versions will be used for linking in preference to the static ones. This means that if we switch to using the default_library option for building only static/shared/both we also need a separate flag for indicating whether to link apps statically or dynamically in the "both" case. I'm also a bit curious as to how much benefit you would expect to come from this change. Since in 99% of cases the C files are only ever built once and then put into the static and shared libs, I would not expect there to be a large build-time improvement from disabling one of the library types. However, having never tested it, I may be wrong on this and there may be a boost from disabling the shared libs - though that in turn means fewer build-time checks for the version.map files, for example. Let me know if you do try it and see a big benefit. Hope this input helps. Regards, /Bruce
Re: [PATCH v4] examples/vm_power_manager: use safe version of list iterator
On 22/08/2022 11:58, Reshma Pattan wrote:
From: Hamza Khan
Currently, when vm_power_manager exits, we are using a LIST_FOREACH
macro to iterate over VM info structures while freeing them. This
leads to use-after-free error. To address this, replace all usages of
LIST_* with TAILQ_* macros, and use the RTE_TAILQ_FOREACH_SAFE macro
to iterate and delete VM info structures.
* The change is small and doesn’t affect other code
* Testing was performed on the patch
Fixes: e8ae9b662506 ("examples/vm_power: channel manager and monitor in host")
Cc: alan.ca...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Hamza Khan
Reviewed-by: Reshma Pattan
Acked-by: Reshma Pattan
Signed-off-by: Reshma Pattan
---
v4: fix header file inclusion
---
--snip--
Acked-by: David Hunt
[PATCH v2] examples/fips_validation: add parsing for AES CTR
Added functionality to parse algorithm for AES CTR test
Signed-off-by: Brian Dooley
---
v2: fix clang warning for int-in-bool-context
---
examples/fips_validation/fips_validation.c | 2 ++
examples/fips_validation/fips_validation.h | 2 ++
examples/fips_validation/fips_validation_aes.c | 5 +
examples/fips_validation/main.c| 9 +++--
4 files changed, 16 insertions(+), 2 deletions(-)
diff --git a/examples/fips_validation/fips_validation.c
b/examples/fips_validation/fips_validation.c
index 12b9b03f56..541eead078 100644
--- a/examples/fips_validation/fips_validation.c
+++ b/examples/fips_validation/fips_validation.c
@@ -466,6 +466,8 @@ fips_test_parse_one_json_vector_set(void)
info.algo = FIPS_TEST_ALGO_AES_CBC;
else if (strstr(algo_str, "AES-XTS"))
info.algo = FIPS_TEST_ALGO_AES_XTS;
+ else if (strstr(algo_str, "AES-CTR"))
+ info.algo = FIPS_TEST_ALGO_AES_CTR;
else if (strstr(algo_str, "SHA"))
info.algo = FIPS_TEST_ALGO_SHA;
else
diff --git a/examples/fips_validation/fips_validation.h
b/examples/fips_validation/fips_validation.h
index 5c1abcbd91..96fdbec41a 100644
--- a/examples/fips_validation/fips_validation.h
+++ b/examples/fips_validation/fips_validation.h
@@ -35,6 +35,7 @@
enum fips_test_algorithms {
FIPS_TEST_ALGO_AES = 0,
FIPS_TEST_ALGO_AES_CBC,
+ FIPS_TEST_ALGO_AES_CTR,
FIPS_TEST_ALGO_AES_GCM,
FIPS_TEST_ALGO_AES_CMAC,
FIPS_TEST_ALGO_AES_CCM,
@@ -105,6 +106,7 @@ enum fips_aesavs_test_types {
AESAVS_TYPE_MMT,
AESAVS_TYPE_MCT,
AESAVS_TYPE_AFT,
+ AESAVS_TYPE_CTR,
};
enum fips_tdes_test_types {
diff --git a/examples/fips_validation/fips_validation_aes.c
b/examples/fips_validation/fips_validation_aes.c
index 4f61505bb3..0ef97aa03d 100644
--- a/examples/fips_validation/fips_validation_aes.c
+++ b/examples/fips_validation/fips_validation_aes.c
@@ -30,8 +30,10 @@
#define TESTTYPE_JSON_STR "testType"
#define DIR_JSON_STR "direction"
#define KEYLEN_JSON_STR"keyLen"
+#define OVERFLOW_JSON_STR "overflow"
#define KEY_JSON_STR "key"
+#define PAYLOADLEN_JSON_STR"payloadLen"
#define IV_JSON_STR"iv"
#define PT_JSON_STR"pt"
#define CT_JSON_STR"ct"
@@ -52,6 +54,7 @@ struct {
{AESAVS_TYPE_MMT, "MMT"},
{AESAVS_TYPE_MCT, "MCT"},
{AESAVS_TYPE_AFT, "AFT"},
+ {AESAVS_TYPE_CTR, "CTR"},
};
struct aes_test_algo {
@@ -60,6 +63,7 @@ struct aes_test_algo {
} const algo_con[] = {
{"CBC", RTE_CRYPTO_CIPHER_AES_CBC},
{"ECB", RTE_CRYPTO_CIPHER_AES_ECB},
+ {"CTR", RTE_CRYPTO_CIPHER_AES_CTR},
};
static int
@@ -291,6 +295,7 @@ parse_test_aes_json_init(void)
case AESAVS_TYPE_MCT:
info.parse_writeback = parse_test_aes_mct_json_writeback;
break;
+ case AESAVS_TYPE_CTR:
case AESAVS_TYPE_AFT:
info.parse_writeback = parse_test_aes_json_writeback;
break;
diff --git a/examples/fips_validation/main.c b/examples/fips_validation/main.c
index 8bd5a66889..0ee618cc66 100644
--- a/examples/fips_validation/main.c
+++ b/examples/fips_validation/main.c
@@ -770,9 +770,11 @@ prepare_aes_xform(struct rte_crypto_sym_xform *xform)
struct rte_crypto_cipher_xform *cipher_xform = &xform->cipher;
xform->type = RTE_CRYPTO_SYM_XFORM_CIPHER;
-
if (info.interim_info.aes_data.cipher_algo == RTE_CRYPTO_CIPHER_AES_CBC)
cipher_xform->algo = RTE_CRYPTO_CIPHER_AES_CBC;
+ else if (info.interim_info.aes_data.cipher_algo ==
+ RTE_CRYPTO_CIPHER_AES_CTR)
+ cipher_xform->algo = RTE_CRYPTO_CIPHER_AES_CTR;
else
cipher_xform->algo = RTE_CRYPTO_CIPHER_AES_ECB;
@@ -781,7 +783,8 @@ prepare_aes_xform(struct rte_crypto_sym_xform *xform)
RTE_CRYPTO_CIPHER_OP_DECRYPT;
cipher_xform->key.data = vec.cipher_auth.key.val;
cipher_xform->key.length = vec.cipher_auth.key.len;
- if (cipher_xform->algo == RTE_CRYPTO_CIPHER_AES_CBC) {
+ if (cipher_xform->algo == RTE_CRYPTO_CIPHER_AES_CBC ||
+ cipher_xform->algo == RTE_CRYPTO_CIPHER_AES_CTR) {
cipher_xform->iv.length = vec.iv.len;
cipher_xform->iv.offset = IV_OFF;
} else {
@@ -1796,6 +1799,7 @@ init_test_ops(void)
{
switch (info.algo) {
case FIPS_TEST_ALGO_AES_CBC:
+ case FIPS_TEST_ALGO_AES_CTR:
case FIPS_TEST_ALGO_AES:
test_ops.prepare_op = prepare_cipher_op;
test_ops.prepare_xform = prepare_aes_xform;
@@ -2007,6 +2011,7 @@ fips_test_one_test_group(void)
ret = parse_test_xts_json_init();
break;
case FIPS_TEST_ALGO
Re: [PATCH v4 2/2] net/cnxk: support ops to update precolor VLAN table
On Tue, Jun 21, 2022 at 1:05 PM wrote:
>
> From: Sunil Kumar Kori
>
> Implement API to update VLAN table for pre-coloring for
> incoming packet per nixlf for CN10K platform.
>
> Signed-off-by: Sunil Kumar Kori
Squashed 1/2 and 2/2 and Updated the git commit as follows and applied
to dpdk-next-net-eventdev/for-main. Thanks
net/cnxk: support for ingress meter pre-color
Added support for ingress meter pre-coloring for incoming
packet for CN10K platform.
Signed-off-by: Sunil Kumar Kori
> ---
> v3..v4:
> - Remove Depends On tag from commit log.
> - Rebase on top of dpdk-next-net/main.
> - Fix clang build failures.
>
> v2..v3:
> - Fix dscp table runtime update error.
>
> v1..v2:
> - Aligned with latest input color spec.
>
> drivers/net/cnxk/cnxk_ethdev.c | 1 +
> drivers/net/cnxk/cnxk_ethdev.h | 3 +-
> drivers/net/cnxk/cnxk_ethdev_mtr.c | 282 ++---
> 3 files changed, 259 insertions(+), 27 deletions(-)
>
> diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c
> index 09e57361d2..55945456c1 100644
> --- a/drivers/net/cnxk/cnxk_ethdev.c
> +++ b/drivers/net/cnxk/cnxk_ethdev.c
> @@ -1679,6 +1679,7 @@ cnxk_eth_dev_init(struct rte_eth_dev *eth_dev)
> dev->eth_dev = eth_dev;
> dev->configured = 0;
> dev->ptype_disable = 0;
> + dev->proto = RTE_MTR_COLOR_IN_PROTO_OUTER_VLAN;
>
> TAILQ_INIT(&dev->inb.list);
> TAILQ_INIT(&dev->outb.list);
> diff --git a/drivers/net/cnxk/cnxk_ethdev.h b/drivers/net/cnxk/cnxk_ethdev.h
> index e99230285c..feb24f2839 100644
> --- a/drivers/net/cnxk/cnxk_ethdev.h
> +++ b/drivers/net/cnxk/cnxk_ethdev.h
> @@ -412,7 +412,8 @@ struct cnxk_eth_dev {
> uint64_t clk_delta;
>
> /* Ingress policer */
> - enum roc_nix_bpf_color precolor_tbl[ROC_NIX_BPF_PRE_COLOR_MAX];
> + enum roc_nix_bpf_color
> precolor_tbl[ROC_NIX_BPF_PRECOLOR_TBL_SIZE_DSCP];
> + enum rte_mtr_color_in_protocol proto;
> struct cnxk_mtr_profiles mtr_profiles;
> struct cnxk_mtr_policy mtr_policy;
> struct cnxk_mtr mtr;
> diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c
> b/drivers/net/cnxk/cnxk_ethdev_mtr.c
> index 02803bdf75..be2cb7d628 100644
> --- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
> +++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
> @@ -48,7 +48,12 @@ static struct rte_mtr_capabilities mtr_capa = {
> RTE_MTR_STATS_N_PKTS_RED | RTE_MTR_STATS_N_PKTS_DROPPED
> |
> RTE_MTR_STATS_N_BYTES_GREEN |
> RTE_MTR_STATS_N_BYTES_YELLOW |
> RTE_MTR_STATS_N_BYTES_RED |
> - RTE_MTR_STATS_N_BYTES_DROPPED};
> + RTE_MTR_STATS_N_BYTES_DROPPED,
> + .input_color_proto_mask = RTE_MTR_COLOR_IN_PROTO_OUTER_VLAN |
> + RTE_MTR_COLOR_IN_PROTO_INNER_VLAN |
> + RTE_MTR_COLOR_IN_PROTO_OUTER_IP |
> + RTE_MTR_COLOR_IN_PROTO_INNER_IP,
> + .separate_input_color_table_per_port = true};
>
> static struct cnxk_meter_node *
> nix_mtr_find(struct cnxk_eth_dev *dev, uint32_t meter_id)
> @@ -470,6 +475,7 @@ cnxk_nix_mtr_create(struct rte_eth_dev *eth_dev, uint32_t
> mtr_id,
> struct cnxk_mtr_profile_node *profile;
> struct cnxk_mtr_policy_node *policy;
> struct cnxk_mtr *fm = &dev->mtr;
> + enum rte_color *table = NULL;
> struct cnxk_meter_node *mtr;
> int i;
>
> @@ -521,18 +527,40 @@ cnxk_nix_mtr_create(struct rte_eth_dev *eth_dev,
> uint32_t mtr_id,
> mtr->is_next = false;
> mtr->level = ROC_NIX_BPF_LEVEL_IDX_INVALID;
>
> + /* populate dscp table for input coloring */
> if (params->dscp_table) {
> - mtr->params.dscp_table =
> - plt_zmalloc(ROC_NIX_BPF_PRE_COLOR_MAX, ROC_ALIGN);
> - if (mtr->params.dscp_table == NULL) {
> + table = (enum rte_color *)plt_zmalloc(sizeof(enum rte_color) *
> + ROC_NIX_BPF_PRECOLOR_TBL_SIZE_DSCP, ROC_ALIGN);
> + if (table == NULL) {
> plt_free(mtr);
> return -rte_mtr_error_set(error, ENOMEM,
> RTE_MTR_ERROR_TYPE_UNSPECIFIED,
> NULL, "Memory alloc failed.");
> }
>
> - for (i = 0; i < ROC_NIX_BPF_PRE_COLOR_MAX; i++)
> - mtr->params.dscp_table[i] = params->dscp_table[i];
> + for (i = 0; i < ROC_NIX_BPF_PRECOLOR_TBL_SIZE_DSCP; i++)
> + table[i] = params->dscp_table[i];
> +
> + mtr->params.dscp_table = table;
> + }
> +
> +
> + /* populate vlan table for input coloring */
> + if (params->vlan_table) {
> + table = (enum rte_color *)plt_zmalloc(sizeof(enum rte_color) *
> + ROC_NIX_BPF_PRECOLOR_TBL_SIZE_VLAN, ROC_ALIGN);
> +
Re: [PATCH 2/3] eal: uninline rte_str_to_size
On Mon, Aug 22, 2022 at 09:24:47AM +0200, Morten Brørup wrote: > > From: Dmitry Kozlyuk [mailto:dmitry.kozl...@gmail.com] > > Sent: Sunday, 21 August 2022 22.50 > > To: dev@dpdk.org > > Cc: Dmitry Kozlyuk; Ray Kinsella > > Subject: [PATCH 2/3] eal: uninline rte_str_to_size > > > > There is no reason for rte_str_to_size() to be inline. > > Move the implementation out of . > > Export it as a stable ABI because it always has been public. > > > > Signed-off-by: Dmitry Kozlyuk > > Acked-by: Morten Brørup > > > --- > > Now doesn't need to #include and , > > but removing them breaks some DPDK code, may break user code too. > > I'm not sure what is the compatibility policy in this regard. > > If such a breakage is allowed, I'd remove includes and fix DPDK code. > > > > The question I'm asking myself here is: Do we want rte_common.h to include > common headers like these, just so we don't need to include them elsewhere? I > think not. > > I'm in favor of the principle of keeping it clean: Remove them from > rte_common.h, and deal with the consequences. > > If we keep them, we will forget why they are there, and some day in the > future, someone will ask what these unused headers are doing in > . > +1 Since removing headers is a build-time issue only and not runtime, I think we should just remove them. /Bruce
Re: [PATCH 0/3] eal: small rte_common.h fixes and cleanup
On Sun, Aug 21, 2022 at 11:50:06PM +0300, Dmitry Kozlyuk wrote: > Dmitry Kozlyuk (3): > eal: fix pointer arithmetic with an expression argument > eal: uninline rte_str_to_size > eal: deduplicate roundup code > > app/test/test_common.c | 11 > lib/eal/common/eal_common_string_fns.c | 32 ++ > lib/eal/include/rte_common.h | 38 -- > lib/eal/version.map| 1 + > 4 files changed, 49 insertions(+), 33 deletions(-) > Thanks. Series-acked-by: Bruce Richardson
[PATCH v3 1/5] mbuf: clarify meta data needed for Outbound Inline
Clarify mbuf meta data needed for Outbound Inline processing.
Application needs to provide mbuf.l3_len and L3 type in
mbuf.ol_flags so that like tunnel mode using mbuf.l2_len, transport mode
can make use of l3_len and l3_type to determine perform
proper transport mode IPsec processing.
Signed-off-by: Nithin Dabilpuram
---
v3:
- Addressed comments on patch 4/5 and added acks.
v2:
- Modified ipsec-secgw to do ether type update for outbound path.
doc/guides/nics/features.rst | 2 +-
lib/mbuf/rte_mbuf_core.h | 3 ++-
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/doc/guides/nics/features.rst b/doc/guides/nics/features.rst
index 7f6cb91..b4a8e98 100644
--- a/doc/guides/nics/features.rst
+++ b/doc/guides/nics/features.rst
@@ -431,7 +431,7 @@ protocol operations. See security library and PMD
documentation for more details
* **[uses] rte_eth_rxconf,rte_eth_rxmode**:
``offloads:RTE_ETH_RX_OFFLOAD_SECURITY``,
* **[uses] rte_eth_txconf,rte_eth_txmode**:
``offloads:RTE_ETH_TX_OFFLOAD_SECURITY``.
-* **[uses] mbuf**: ``mbuf.l2_len``.
+* **[uses] mbuf**: ``mbuf.l2_len``, ``mbuf.l3_len``, ``mbuf.ol_flags``.
* **[implements] rte_security_ops**: ``session_create``, ``session_update``,
``session_stats_get``, ``session_destroy``, ``set_pkt_metadata``,
``get_userdata``,
``capabilities_get``.
diff --git a/lib/mbuf/rte_mbuf_core.h b/lib/mbuf/rte_mbuf_core.h
index 3d6ddd6..b62a7c6 100644
--- a/lib/mbuf/rte_mbuf_core.h
+++ b/lib/mbuf/rte_mbuf_core.h
@@ -267,7 +267,8 @@ extern "C" {
/**
* Request security offload processing on the TX packet.
* To use Tx security offload, the user needs to fill l2_len in mbuf
- * indicating L2 header size and where L3 header starts.
+ * indicating L2 header size and where L3 header starts. Similarly,
+ * l3_len should also be filled along with ol_flags reflecting current L3 type.
*/
#define RTE_MBUF_F_TX_SEC_OFFLOAD (1ULL << 43)
#define PKT_TX_SEC_OFFLOAD RTE_DEPRECATED(PKT_TX_SEC_OFFLOAD) \
--
2.8.4
[PATCH v3 2/5] security: clarify L2 header requirement for outbound inline
Clarify that for Outbound Inline IPsec processing, L2 header needs to be up to date with ether type which will be applicable post IPsec processing as the IPsec offload only touches L3 and above. Signed-off-by: Nithin Dabilpuram --- doc/guides/prog_guide/rte_security.rst | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/doc/guides/prog_guide/rte_security.rst b/doc/guides/prog_guide/rte_security.rst index 72ca0bd..be158f6 100644 --- a/doc/guides/prog_guide/rte_security.rst +++ b/doc/guides/prog_guide/rte_security.rst @@ -146,7 +146,9 @@ adding the relevant protocol headers and encrypting the data before sending the packet out. The software should make sure that the buffer has required head room and tail room for any protocol header addition. The software may also do early fragmentation if the resultant packet is expected -to cross the MTU size. +to cross the MTU size. The software should also make sure that L2 header contents +are updated with the final L2 header which is expected post IPsec processing as +the IPsec offload will only update L3 and above in egress path. .. note:: -- 2.8.4
[PATCH v3 3/5] net/cnxk: remove L2 header update for outbound inline pkts
Remove L2 header update for outbound inline packets as
application is already taking care of the same.
Signed-off-by: Nithin Dabilpuram
---
drivers/net/cnxk/cn10k_tx.h | 17 -
1 file changed, 17 deletions(-)
diff --git a/drivers/net/cnxk/cn10k_tx.h b/drivers/net/cnxk/cn10k_tx.h
index ea13866..4bd47ef 100644
--- a/drivers/net/cnxk/cn10k_tx.h
+++ b/drivers/net/cnxk/cn10k_tx.h
@@ -362,15 +362,6 @@ cn10k_nix_prep_sec_vec(struct rte_mbuf *m, uint64x2_t
*cmd0, uint64x2_t *cmd1,
dptr += l2_len;
- if (sess_priv.mode == ROC_IE_SA_MODE_TUNNEL) {
- if (sess_priv.outer_ip_ver == ROC_IE_SA_IP_VERSION_4)
- *((uint16_t *)(dptr - 2)) =
- rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
- else
- *((uint16_t *)(dptr - 2)) =
- rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
- }
-
ucode_cmd[1] = dptr;
ucode_cmd[2] = dptr;
@@ -490,14 +481,6 @@ cn10k_nix_prep_sec(struct rte_mbuf *m, uint64_t *cmd,
uintptr_t *nixtx_addr,
dptr += l2_len;
- if (sess_priv.mode == ROC_IE_SA_MODE_TUNNEL) {
- if (sess_priv.outer_ip_ver == ROC_IE_SA_IP_VERSION_4)
- *((uint16_t *)(dptr - 2)) =
- rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
- else
- *((uint16_t *)(dptr - 2)) =
- rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
- }
ucode_cmd[1] = dptr;
ucode_cmd[2] = dptr;
--
2.8.4
[PATCH v3 4/5] app/test: update L2 header based on tunnel IP version
Update L2 header based on tunnel IP version in the application
as driver/HW is not expected to update L2 ether type post
Outbound Inline protocol offload processing.
Signed-off-by: Nithin Dabilpuram
Acked-by: Akhil Goyal
---
app/test/test_security_inline_proto.c | 34 +-
1 file changed, 29 insertions(+), 5 deletions(-)
diff --git a/app/test/test_security_inline_proto.c
b/app/test/test_security_inline_proto.c
index 5f26a04..b282e7d 100644
--- a/app/test/test_security_inline_proto.c
+++ b/app/test/test_security_inline_proto.c
@@ -418,15 +418,29 @@ copy_buf_to_pkt_segs(const uint8_t *buf, unsigned int len,
rte_memcpy(seg_buf, buf + copied, (size_t) len);
}
+static bool
+is_outer_ipv4(struct ipsec_test_data *td)
+{
+ bool outer_ipv4;
+
+ if (td->ipsec_xform.direction == RTE_SECURITY_IPSEC_SA_DIR_INGRESS ||
+ td->ipsec_xform.mode == RTE_SECURITY_IPSEC_SA_MODE_TRANSPORT)
+ outer_ipv4 = (((td->input_text.data[0] & 0xF0) >> 4) ==
IPVERSION);
+ else
+ outer_ipv4 = (td->ipsec_xform.tunnel.type ==
RTE_SECURITY_IPSEC_TUNNEL_IPV4);
+ return outer_ipv4;
+}
+
static inline struct rte_mbuf *
-init_packet(struct rte_mempool *mp, const uint8_t *data, unsigned int len)
+init_packet(struct rte_mempool *mp, const uint8_t *data, unsigned int len,
bool outer_ipv4)
{
struct rte_mbuf *pkt;
pkt = rte_pktmbuf_alloc(mp);
if (pkt == NULL)
return NULL;
- if (((data[0] & 0xF0) >> 4) == IPVERSION) {
+
+ if (outer_ipv4) {
rte_memcpy(rte_pktmbuf_append(pkt, RTE_ETHER_HDR_LEN),
&dummy_ipv4_eth_hdr, RTE_ETHER_HDR_LEN);
pkt->l3_len = sizeof(struct rte_ipv4_hdr);
@@ -711,6 +725,7 @@ test_ipsec_with_reassembly(struct reassembly_vector *vector,
struct rte_security_ctx *ctx;
unsigned int i, nb_rx = 0, j;
uint32_t ol_flags;
+ bool outer_ipv4;
int ret = 0;
burst_sz = vector->burst ? ENCAP_DECAP_BURST_SZ : 1;
@@ -740,11 +755,15 @@ test_ipsec_with_reassembly(struct reassembly_vector
*vector,
memset(tx_pkts_burst, 0, sizeof(tx_pkts_burst[0]) * nb_tx);
memset(rx_pkts_burst, 0, sizeof(rx_pkts_burst[0]) * nb_tx);
+ memcpy(&sa_data, vector->sa_data, sizeof(struct ipsec_test_data));
+ sa_data.ipsec_xform.direction = RTE_SECURITY_IPSEC_SA_DIR_EGRESS;
+ outer_ipv4 = is_outer_ipv4(&sa_data);
+
for (i = 0; i < nb_tx; i += vector->nb_frags) {
for (j = 0; j < vector->nb_frags; j++) {
tx_pkts_burst[i+j] = init_packet(mbufpool,
vector->frags[j]->data,
- vector->frags[j]->len);
+ vector->frags[j]->len,
outer_ipv4);
if (tx_pkts_burst[i+j] == NULL) {
ret = -1;
printf("\n packed init failed\n");
@@ -963,6 +982,7 @@ test_ipsec_inline_proto_process(struct ipsec_test_data *td,
int nb_rx = 0, nb_sent;
uint32_t ol_flags;
int i, j = 0, ret;
+ bool outer_ipv4;
memset(rx_pkts_burst, 0, sizeof(rx_pkts_burst[0]) * nb_pkts);
@@ -994,9 +1014,11 @@ test_ipsec_inline_proto_process(struct ipsec_test_data
*td,
if (ret)
goto out;
}
+ outer_ipv4 = is_outer_ipv4(td);
+
for (i = 0; i < nb_pkts; i++) {
tx_pkts_burst[i] = init_packet(mbufpool, td->input_text.data,
- td->input_text.len);
+ td->input_text.len, outer_ipv4);
if (tx_pkts_burst[i] == NULL) {
while (i--)
rte_pktmbuf_free(tx_pkts_burst[i]);
@@ -1194,6 +1216,7 @@ test_ipsec_inline_proto_process_with_esn(struct
ipsec_test_data td[],
struct rte_security_session *ses;
struct rte_security_ctx *ctx;
uint32_t ol_flags;
+ bool outer_ipv4;
int i, ret;
if (td[0].aead) {
@@ -1224,10 +1247,11 @@ test_ipsec_inline_proto_process_with_esn(struct
ipsec_test_data td[],
if (ret)
goto out;
}
+ outer_ipv4 = is_outer_ipv4(td);
for (i = 0; i < nb_pkts; i++) {
tx_pkt = init_packet(mbufpool, td[i].input_text.data,
- td[i].input_text.len);
+ td[i].input_text.len, outer_ipv4);
if (tx_pkt == NULL) {
ret = TEST_FAILED;
goto out;
--
2.8.4
[PATCH v3 5/5] examples/ipsec-secgw: update ether type using tunnel info
Update ether type for outbound SA processing based on tunnel header
information in both NEON functions for poll mode and event mode worker
functions.
Signed-off-by: Nithin Dabilpuram
Reviewed-by: Ruifeng Wang
Acked-by: Akhil Goyal
---
examples/ipsec-secgw/ipsec_neon.h | 41 +
examples/ipsec-secgw/ipsec_worker.c | 30 +++
2 files changed, 49 insertions(+), 22 deletions(-)
diff --git a/examples/ipsec-secgw/ipsec_neon.h
b/examples/ipsec-secgw/ipsec_neon.h
index 3f2d0a0..9c0498b 100644
--- a/examples/ipsec-secgw/ipsec_neon.h
+++ b/examples/ipsec-secgw/ipsec_neon.h
@@ -18,12 +18,13 @@ extern xmm_t val_eth[RTE_MAX_ETHPORTS];
*/
static inline void
processx4_step3(struct rte_mbuf *pkts[FWDSTEP], uint16_t dst_port[FWDSTEP],
- uint64_t tx_offloads, bool ip_cksum, uint8_t *l_pkt)
+ uint64_t tx_offloads, bool ip_cksum, bool is_ipv4, uint8_t
*l_pkt)
{
uint32x4_t te[FWDSTEP];
uint32x4_t ve[FWDSTEP];
uint32_t *p[FWDSTEP];
struct rte_mbuf *pkt;
+ uint32_t val;
uint8_t i;
for (i = 0; i < FWDSTEP; i++) {
@@ -38,7 +39,15 @@ processx4_step3(struct rte_mbuf *pkts[FWDSTEP], uint16_t
dst_port[FWDSTEP],
te[i] = vld1q_u32(p[i]);
/* Update last 4 bytes */
- ve[i] = vsetq_lane_u32(vgetq_lane_u32(te[i], 3), ve[i], 3);
+ val = vgetq_lane_u32(te[i], 3);
+#if RTE_BYTE_ORDER == RTE_LITTLE_ENDIAN
+ val &= 0xUL << 16;
+ val |= rte_cpu_to_be_16(is_ipv4 ? RTE_ETHER_TYPE_IPV4 :
RTE_ETHER_TYPE_IPV6);
+#else
+ val &= 0xUL;
+ val |= rte_cpu_to_be_16(is_ipv4 ? RTE_ETHER_TYPE_IPV4 :
RTE_ETHER_TYPE_IPV6) << 16;
+#endif
+ ve[i] = vsetq_lane_u32(val, ve[i], 3);
vst1q_u32(p[i], ve[i]);
if (ip_cksum) {
@@ -64,10 +73,11 @@ processx4_step3(struct rte_mbuf *pkts[FWDSTEP], uint16_t
dst_port[FWDSTEP],
*/
static inline void
process_packet(struct rte_mbuf *pkt, uint16_t *dst_port, uint64_t tx_offloads,
- bool ip_cksum, uint8_t *l_pkt)
+ bool ip_cksum, bool is_ipv4, uint8_t *l_pkt)
{
struct rte_ether_hdr *eth_hdr;
uint32x4_t te, ve;
+ uint32_t val;
/* Check if it is a large packet */
if (pkt->pkt_len - RTE_ETHER_HDR_LEN > mtu_size)
@@ -78,7 +88,15 @@ process_packet(struct rte_mbuf *pkt, uint16_t *dst_port,
uint64_t tx_offloads,
te = vld1q_u32((uint32_t *)eth_hdr);
ve = vreinterpretq_u32_s32(val_eth[dst_port[0]]);
- ve = vcopyq_laneq_u32(ve, 3, te, 3);
+ val = vgetq_lane_u32(te, 3);
+#if RTE_BYTE_ORDER == RTE_LITTLE_ENDIAN
+ val &= 0xUL << 16;
+ val |= rte_cpu_to_be_16(is_ipv4 ? RTE_ETHER_TYPE_IPV4 :
RTE_ETHER_TYPE_IPV6);
+#else
+ val &= 0xUL;
+ val |= rte_cpu_to_be_16(is_ipv4 ? RTE_ETHER_TYPE_IPV4 :
RTE_ETHER_TYPE_IPV6) << 16;
+#endif
+ ve = vsetq_lane_u32(val, ve, 3);
vst1q_u32((uint32_t *)eth_hdr, ve);
if (ip_cksum) {
@@ -223,14 +241,14 @@ send_multi_pkts(struct rte_mbuf **pkts, uint16_t
dst_port[MAX_PKT_BURST],
lp = pnum;
lp[0] = 1;
- processx4_step3(pkts, dst_port, tx_offloads, ip_cksum, &l_pkt);
+ processx4_step3(pkts, dst_port, tx_offloads, ip_cksum, is_ipv4,
&l_pkt);
/* dp1: */
dp1 = vld1q_u16(dst_port);
for (i = FWDSTEP; i != k; i += FWDSTEP) {
- processx4_step3(&pkts[i], &dst_port[i], tx_offloads,
- ip_cksum, &l_pkt);
+ processx4_step3(&pkts[i], &dst_port[i], tx_offloads,
ip_cksum, is_ipv4,
+ &l_pkt);
/*
* dp2:
@@ -268,20 +286,17 @@ send_multi_pkts(struct rte_mbuf **pkts, uint16_t
dst_port[MAX_PKT_BURST],
/* Process up to last 3 packets one by one. */
switch (nb_rx % FWDSTEP) {
case 3:
- process_packet(pkts[i], dst_port + i, tx_offloads, ip_cksum,
- &l_pkt);
+ process_packet(pkts[i], dst_port + i, tx_offloads, ip_cksum,
is_ipv4, &l_pkt);
GROUP_PORT_STEP(dlp, dst_port, lp, pnum, i);
i++;
/* fallthrough */
case 2:
- process_packet(pkts[i], dst_port + i, tx_offloads, ip_cksum,
- &l_pkt);
+ process_packet(pkts[i], dst_port + i, tx_offloads, ip_cksum,
is_ipv4, &l_pkt);
GROUP_PORT_STEP(dlp, dst_port, lp, pnum, i);
i++;
/* fallthrough */
case 1:
- process_packet(pkts[i], dst_port + i, tx_offloads, ip_cksum,
- &l_pkt);
+ process_packet(pkts[i], dst_port + i, tx_offloads, ip_cksum,
is
Re: [Patch v4 00/17] Introduce Microsoft Azure Network Adatper (MANA) PMD
On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote: From: Long Li MANA is a network interface card to be used in the Azure cloud environment. MANA provides safe access to user memory through memory registration. It has IOMMU built into the hardware. MANA uses IB verbs and RDMA layer to configure hardware resources. It requires the corresponding RDMA kernel-mode and user-mode drivers. The MANA RDMA kernel-mode driver is being reviewed at: https://patchwork.kernel.org/project/netdevbpf/cover/1655345240-26411-1-git-send-email-lon...@linuxonhyperv.com/ The MANA RDMA user-mode driver is being reviewed at: https://github.com/linux-rdma/rdma-core/pull/1177 Hi Long, How are the dependencies are progressing? Is there anything blocker for DPDK upstream? Long Li (17): net/mana: add basic driver, build environment and doc net/mana: add device configuration and stop net/mana: add function to report support ptypes net/mana: add link update net/mana: add function for device removal interrupts net/mana: add device info net/mana: add function to configure RSS net/mana: add function to configure RX queues net/mana: add function to configure TX queues net/mana: implement memory registration net/mana: implement the hardware layer operations net/mana: add function to start/stop TX queues net/mana: add function to start/stop RX queues net/mana: add function to receive packets net/mana: add function to send packets net/mana: add function to start/stop device net/mana: add function to report queue stats <...>
Re: [Patch v4 01/17] net/mana: add basic driver, build environment and doc
On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote:
CAUTION: This message has originated from an External Source. Please use proper
judgment and caution when opening attachments, clicking links, or responding to
this email.
From: Long Li
MANA is a PCI device. It uses IB verbs to access hardware through the
kernel RDMA layer. This patch introduces build environment and basic
device probe functions.
Signed-off-by: Long Li
---
Change log:
v2:
Fix typos.
Make the driver build only on x86-64 and Linux.
Remove unused header files.
Change port definition to uint16_t or uint8_t (for IB).
Use getline() in place of fgets() to read and truncate a line.
v3:
Add meson build check for required functions from RDMA direct verb header file
v4:
Remove extra "\n" in logging code.
Use "r" in place of "rb" in fopen() to read text files.
<...>
--- /dev/null
+++ b/doc/guides/nics/mana.rst
@@ -0,0 +1,66 @@
+.. SPDX-License-Identifier: BSD-3-Clause
+Copyright 2022 Microsoft Corporation
+
+MANA poll mode driver library
+=
+
+The MANA poll mode driver library (**librte_net_mana**) implements support
+for Microsoft Azure Network Adapter VF in SR-IOV context.
+
Can you please provide any link to an official product description? As a
reference point for anybody interested more with the product details.
<..>
+
+Netvsc PMD arguments > +
'Netvsc'? Do you mean 'MANA'?
j
+
+The user can specify below argument in devargs.
+
+#. ``mac``:
+
+Specify the MAC address for this device. If it is set, the driver
+probes and loads the NIC with a matching mac address. If it is not
+set, the driver probes on all the NICs on the PCI device. The default
+value is not set, meaning all the NICs will be probed and loaded.
Code accepts up to 8 mac value, should this be documented?
Also why this devarg is needed?
diff --git a/drivers/net/mana/mana.c b/drivers/net/mana/mana.c
new file mode 100644
index 00..cb59eb6882
--- /dev/null
+++ b/drivers/net/mana/mana.c
@@ -0,0 +1,704 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright 2022 Microsoft Corporation
+ */
+
+#include
+#include
+#include
+#include
+
+#include
+#include
+#include
+#include
+
+#include
+#include
+
+#include
+
+#include "mana.h"
+
+/* Shared memory between primary/secondary processes, per driver */
+struct mana_shared_data *mana_shared_data;
+const struct rte_memzone *mana_shared_mz;
If these global variables are not used by other compilation units,
please try to make them static as much as possible.
+static const char *MZ_MANA_SHARED_DATA = "mana_shared_data";
+
+struct mana_shared_data mana_local_data;
+
Can you put some comment to this global variables?
+/* Spinlock for mana_shared_data */
+static rte_spinlock_t mana_shared_data_lock = RTE_SPINLOCK_INITIALIZER;
+
+/* Allocate a buffer on the stack and fill it with a printf format string. */
+#define MKSTR(name, ...) \
+ int mkstr_size_##name = snprintf(NULL, 0, "" __VA_ARGS__); \
+ char name[mkstr_size_##name + 1]; \
+ \
+ memset(name, 0, mkstr_size_##name + 1); \
+ snprintf(name, sizeof(name), "" __VA_ARGS__)
+
+int mana_logtype_driver;
+int mana_logtype_init;
+
+const struct eth_dev_ops mana_dev_ops = {
+};
+
+const struct eth_dev_ops mana_dev_sec_ops = {
+};
It may be better to expand 'sec' to secondary to not confuse with
security etc...
+
+uint16_t
+mana_rx_burst_removed(void *dpdk_rxq __rte_unused,
+ struct rte_mbuf **pkts __rte_unused,
+ uint16_t pkts_n __rte_unused)
+{
+ rte_mb();
+ return 0;
+}
+
+uint16_t
+mana_tx_burst_removed(void *dpdk_rxq __rte_unused,
+ struct rte_mbuf **pkts __rte_unused,
+ uint16_t pkts_n __rte_unused)
+{
+ rte_mb();
+ return 0;
+}
+
+static const char *mana_init_args[] = {
+ "mac",
+ NULL,
+};
+
+/* Support of parsing up to 8 mac address from EAL command line */
+#define MAX_NUM_ADDRESS 8
+struct mana_conf {
+ struct rte_ether_addr mac_array[MAX_NUM_ADDRESS];
+ unsigned int index;
+};
+
+static int mana_arg_parse_callback(const char *key, const char *val,
+ void *private)
Since this is new driver, better to follow the coding convention:
https://doc.dpdk.org/guides/contributing/coding_style.html
Please put return type to another line:
static int
mana_arg_parse_callback(const char *key, const char *val, void *private)
+{
+ struct mana_conf *conf = (struct mana_conf *)private;
+ int ret;
+
+ DRV_LOG(INFO, "key=%s value=%s index=%d", key, val, conf->index);
+
+ if (conf->index >= MAX_NUM_ADDRESS) {
+ DRV_LOG(ERR, "Exceeding max MAC address");
+ return 1;
+ }
+
+ ret = rte_ether_unformat_addr(val, &conf->mac_array[conf->index]);
+ if (ret) {
+ DRV_LOG(ERR, "Invalid MAC address %s", val);
+
Re: [Patch v4 01/17] net/mana: add basic driver, build environment and doc
On 8/22/2022 4:03 PM, Ferruh Yigit wrote:
+ struct rte_pci_device *pci_dev,
+ struct rte_ether_addr *mac_addr)
+{
+ struct ibv_device **ibv_list;
+ int ibv_idx;
+ struct ibv_context *ctx;
+ struct ibv_device_attr_ex dev_attr;
+ int num_devices;
+ int ret = 0;
+ uint8_t port;
+ struct mana_priv *priv = NULL;
+ struct rte_eth_dev *eth_dev = NULL;
+ bool found_port;
+
+ ibv_list = ibv_get_device_list(&num_devices);
+ for (ibv_idx = 0; ibv_idx < num_devices; ibv_idx++) {
+ struct ibv_device *ibdev = ibv_list[ibv_idx];
+ struct rte_pci_addr pci_addr;
+
+ DRV_LOG(INFO, "Probe device name %s dev_name %s
ibdev_path %s",
+ ibdev->name, ibdev->dev_name, ibdev->ibdev_path);
+
+ if (mana_ibv_device_to_pci_addr(ibdev, &pci_addr))
+ continue;
+
+ /* Ignore if this IB device is not this PCI device */
+ if (pci_dev->addr.domain != pci_addr.domain ||
+ pci_dev->addr.bus != pci_addr.bus ||
+ pci_dev->addr.devid != pci_addr.devid ||
+ pci_dev->addr.function != pci_addr.function)
+ continue;
+
As far as I understand, intention of this loop is to find 'ibdev'
matching this device, code gooes through all "ibv device list" for this,
I wonder if there is a easy way for doing this, like a sysfs entry to
help getting this information?
And how mlx4/5 does this?
Since there are multiple RDMA devices now, does it make sense to have
RDMA bus driver, which can hide some PCIe details under bus, and driver
can get PCI and ibdev information during probe?
Re: [Patch v4 11/17] net/mana: implement the hardware layer operations
On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote: From: Long Li The hardware layer of MANA understands the device queue and doorbell formats. Those functions are implemented for use by packet RX/TX code. Signed-off-by: Long Li --- Change log: v2: Remove unused header files. Rename a camel case. <...> +/* NDIS HASH Types */ +#define BIT(nr)(1 << (nr)) +#define NDIS_HASH_IPV4 BIT(0) +#define NDIS_HASH_TCP_IPV4 BIT(1) +#define NDIS_HASH_UDP_IPV4 BIT(2) +#define NDIS_HASH_IPV6 BIT(3) +#define NDIS_HASH_TCP_IPV6 BIT(4) +#define NDIS_HASH_UDP_IPV6 BIT(5) +#define NDIS_HASH_IPV6_EX BIT(6) +#define NDIS_HASH_TCP_IPV6_EX BIT(7) +#define NDIS_HASH_UDP_IPV6_EX BIT(8) Can use RTE_BIT32/64 instead of defining new macro
Re: [Patch v4 17/17] net/mana: add function to report queue stats
On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote:
CAUTION: This message has originated from an External Source. Please use proper
judgment and caution when opening attachments, clicking links, or responding to
this email.
From: Long Li
Report packet statistics.
Signed-off-by: Long Li
<...>
+static int mana_dev_stats_get(struct rte_eth_dev *dev,
+ struct rte_eth_stats *stats)
+{
+ unsigned int i;
+
+ for (i = 0; i < dev->data->nb_tx_queues; i++) {
+ struct mana_txq *txq = dev->data->tx_queues[i];
+
+ if (!txq)
+ continue;
+
+ stats->opackets = txq->stats.packets;
+ stats->obytes = txq->stats.bytes;
+ stats->oerrors = txq->stats.errors;
Shouldn't these be "+=" ? Same for below Rx block.
Re: [Patch v4 15/17] net/mana: add function to send packets
On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote:
CAUTION: This message has originated from an External Source. Please use proper
judgment and caution when opening attachments, clicking links, or responding to
this email.
From: Long Li
With all the TX queues created, MANA can send packets over those queues.
Signed-off-by: Long Li
---
Change log:
v2:
Rename all camel cases.
<...>
+
+ DRV_LOG(DEBUG, "pkt[%d]: buf_addr 0x%p, nb_segs %d, pkt_len %d",
+ pkt_idx, m_pkt->buf_addr, m_pkt->nb_segs,
+ m_pkt->pkt_len);
+
+ /* Create SGL for packet data buffers */
+ for (uint16_t seg_idx = 0; seg_idx < m_pkt->nb_segs; seg_idx++)
{
This is C99 feature and we tend to not use it BUT,
Thomas, David,
In the past there were c89/c90/ansi code in DPDK, but as far as I can
see no more left.
Is there anything that prevents us using C99 features, like variable in
the for loop, I think it is usefull that it reduces the scope and makes
code more readable, etc...
Re: [PATCH] mbuf: remove deprecated offload flags
On Fri, Aug 5, 2022 at 10:03 AM Andrew Rybchenko wrote: > > Remove deprecated ``PKT_*`` flags. Use corresponding flags with > ``RTE_MBUF_F_`` prefix instead. > > Signed-off-by: Andrew Rybchenko The change lgtm. We may reference the cocci script devtools/cocci/prefix_mbuf_offload_flags.cocci in the release notes for people wanting to migrate easily. Wdyt? -- David Marchand
RE: [Patch v4 00/17] Introduce Microsoft Azure Network Adatper (MANA) PMD
> Subject: Re: [Patch v4 00/17] Introduce Microsoft Azure Network Adatper > (MANA) PMD > > On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote: > > > > > From: Long Li > > > > MANA is a network interface card to be used in the Azure cloud > environment. > > MANA provides safe access to user memory through memory registration. > > It has IOMMU built into the hardware. > > > > MANA uses IB verbs and RDMA layer to configure hardware resources. It > > requires the corresponding RDMA kernel-mode and user-mode drivers. > > > > The MANA RDMA kernel-mode driver is being reviewed at: > > > https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fpatc > > hwork.kernel.org%2Fproject%2Fnetdevbpf%2Fcover%2F1655345240- > 26411-1-gi > > t-send-email- > longli%40linuxonhyperv.com%2F&data=05%7C01%7Clongli%4 > > > 0microsoft.com%7C232a4a7af70f490b94f608da844ee782%7C72f988bf86f141a > f91 > > > ab2d7cd011db47%7C1%7C0%7C637967771686870922%7CUnknown%7CTWFp > bGZsb3d8ey > > > JWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D% > 7C300 > > > 0%7C%7C%7C&sdata=bC7l7wxfU5sbahBye8CKVS%2BJvirO%2FvIMU3M > oS4%2Fdpc8 > > %3D&reserved=0 > > > > The MANA RDMA user-mode driver is being reviewed at: > > > https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgith > > ub.com%2Flinux-rdma%2Frdma- > core%2Fpull%2F1177&data=05%7C01%7Clongl > > > i%40microsoft.com%7C232a4a7af70f490b94f608da844ee782%7C72f988bf86f1 > 41a > > > f91ab2d7cd011db47%7C1%7C0%7C637967771686870922%7CUnknown%7CT > WFpbGZsb3d > > > 8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0% > 3D%7C > > > 3000%7C%7C%7C&sdata=DLGg1q1N3%2FNtr7ii3WI2rqShPNoi2u0MF8J8 > Cn1h60s% > > 3D&reserved=0 > > > > Hi Long, > > How are the dependencies are progressing? Is there anything blocker for > DPDK upstream? Hi Ferruh, The kernel and rdma-core user-mode patch submissions are in progress. They are currently at v4. Long > > > > Long Li (17): > >net/mana: add basic driver, build environment and doc > >net/mana: add device configuration and stop > >net/mana: add function to report support ptypes > >net/mana: add link update > >net/mana: add function for device removal interrupts > >net/mana: add device info > >net/mana: add function to configure RSS > >net/mana: add function to configure RX queues > >net/mana: add function to configure TX queues > >net/mana: implement memory registration > >net/mana: implement the hardware layer operations > >net/mana: add function to start/stop TX queues > >net/mana: add function to start/stop RX queues > >net/mana: add function to receive packets > >net/mana: add function to send packets > >net/mana: add function to start/stop device > >net/mana: add function to report queue stats > > > > <...>
RE: [Patch v4 01/17] net/mana: add basic driver, build environment and doc
> Subject: Re: [Patch v4 01/17] net/mana: add basic driver, build environment
> and doc
>
> On 8/22/2022 4:03 PM, Ferruh Yigit wrote:
> >> + struct rte_pci_device *pci_dev,
> >> + struct rte_ether_addr *mac_addr) {
> >> + struct ibv_device **ibv_list;
> >> + int ibv_idx;
> >> + struct ibv_context *ctx;
> >> + struct ibv_device_attr_ex dev_attr;
> >> + int num_devices;
> >> + int ret = 0;
> >> + uint8_t port;
> >> + struct mana_priv *priv = NULL;
> >> + struct rte_eth_dev *eth_dev = NULL;
> >> + bool found_port;
> >> +
> >> + ibv_list = ibv_get_device_list(&num_devices);
> >> + for (ibv_idx = 0; ibv_idx < num_devices; ibv_idx++) {
> >> + struct ibv_device *ibdev = ibv_list[ibv_idx];
> >> + struct rte_pci_addr pci_addr;
> >> +
> >> + DRV_LOG(INFO, "Probe device name %s dev_name %s
> >> ibdev_path %s",
> >> + ibdev->name, ibdev->dev_name,
> >> +ibdev->ibdev_path);
> >> +
> >> + if (mana_ibv_device_to_pci_addr(ibdev, &pci_addr))
> >> + continue;
> >> +
> >> + /* Ignore if this IB device is not this PCI device */
> >> + if (pci_dev->addr.domain != pci_addr.domain ||
> >> + pci_dev->addr.bus != pci_addr.bus ||
> >> + pci_dev->addr.devid != pci_addr.devid ||
> >> + pci_dev->addr.function != pci_addr.function)
> >> + continue;
> >> +
> >
> > As far as I understand, intention of this loop is to find 'ibdev'
> > matching this device, code gooes through all "ibv device list" for
> > this, I wonder if there is a easy way for doing this, like a sysfs
> > entry to help getting this information?
> > And how mlx4/5 does this?
>
> Since there are multiple RDMA devices now, does it make sense to have
> RDMA bus driver, which can hide some PCIe details under bus, and driver can
> get PCI and ibdev information during probe?
Mellanox drivers use a similar way to go through the list of IB devices.
Matan, Viacheslav, what are your thoughts on implementing a bus for RDMA
devices?
RE: [Patch v4 11/17] net/mana: implement the hardware layer operations
> Subject: Re: [Patch v4 11/17] net/mana: implement the hardware layer > operations > > On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote: > > > > > From: Long Li > > > > The hardware layer of MANA understands the device queue and doorbell > > formats. Those functions are implemented for use by packet RX/TX code. > > > > Signed-off-by: Long Li > > --- > > Change log: > > v2: > > Remove unused header files. > > Rename a camel case. > > > > <...> > > > +/* NDIS HASH Types */ > > +#define BIT(nr)(1 << (nr)) > > +#define NDIS_HASH_IPV4 BIT(0) > > +#define NDIS_HASH_TCP_IPV4 BIT(1) > > +#define NDIS_HASH_UDP_IPV4 BIT(2) > > +#define NDIS_HASH_IPV6 BIT(3) > > +#define NDIS_HASH_TCP_IPV6 BIT(4) > > +#define NDIS_HASH_UDP_IPV6 BIT(5) > > +#define NDIS_HASH_IPV6_EX BIT(6) > > +#define NDIS_HASH_TCP_IPV6_EX BIT(7) > > +#define NDIS_HASH_UDP_IPV6_EX BIT(8) > > Can use RTE_BIT32/64 instead of defining new macro Will fix this.
RE: [Patch v4 17/17] net/mana: add function to report queue stats
> Subject: Re: [Patch v4 17/17] net/mana: add function to report queue stats
>
> On 7/9/2022 12:49 AM, lon...@linuxonhyperv.com wrote:
> > CAUTION: This message has originated from an External Source. Please use
> proper judgment and caution when opening attachments, clicking links, or
> responding to this email.
> >
> >
> > From: Long Li
> >
> > Report packet statistics.
> >
> > Signed-off-by: Long Li
>
> <...>
>
> > +static int mana_dev_stats_get(struct rte_eth_dev *dev,
> > + struct rte_eth_stats *stats) {
> > + unsigned int i;
> > +
> > + for (i = 0; i < dev->data->nb_tx_queues; i++) {
> > + struct mana_txq *txq = dev->data->tx_queues[i];
> > +
> > + if (!txq)
> > + continue;
> > +
> > + stats->opackets = txq->stats.packets;
> > + stats->obytes = txq->stats.bytes;
> > + stats->oerrors = txq->stats.errors;
>
>
> Shouldn't these be "+=" ? Same for below Rx block.
Thank you. I will fix those.
[PATCH v2] examples/eventdev_producer_consumer: fix 32-bit checkpatch issues
Fixed style and format issues, primarily those involving data types
whose size varies depending on whether we are building for 32 or
64 bit platforms.
Signed-off-by: Timothy McDaniel
---
examples/eventdev_producer_consumer/main.c | 18 +-
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/examples/eventdev_producer_consumer/main.c
b/examples/eventdev_producer_consumer/main.c
index 54a550b459..164bdf6f74 100644
--- a/examples/eventdev_producer_consumer/main.c
+++ b/examples/eventdev_producer_consumer/main.c
@@ -21,7 +21,7 @@
static unsigned int num_workers = 4;
static bool g_is_mbuf;
-static unsigned long num_packets = (1L << 25); /* do ~32M packets */
+static uint64_t num_packets = (1L << 25); /* do ~32M packets */
static int sched_type = RTE_SCHED_TYPE_ATOMIC;
struct prod_data {
@@ -51,13 +51,13 @@ static struct rte_mempool *mp;
static int
worker(void *arg)
{
- struct rte_event rcv_events[BATCH_SIZE] = {0};
+ struct rte_event rcv_events[BATCH_SIZE];
struct worker_data *data = (struct worker_data *)arg;
uint8_t event_dev_id = data->event_dev_id;
uint8_t event_port_id = data->event_port_id;
int32_t qid = data->qid;
- size_t sent = 0, received = 0;
+ uint64_t sent = 0, received = 0;
uint16_t n;
if (!quiet)
@@ -83,7 +83,7 @@ worker(void *arg)
rte_pause();
continue;
} else if (!quiet)
- printf("Worker received %d events (%zu total)\n",
+ printf("Worker received %d events(%"PRIu64" total)\n",
n, received);
delay_start = rte_rdtsc();
@@ -113,7 +113,7 @@ worker(void *arg)
} /* while (!done) */
if (!quiet)
- printf("%s %d thread done. RX=%zu TX=%zu\n",
+ printf("%s %d thread done. RX= %"PRIu64" TX= %"PRIu64"\n",
__func__, rte_lcore_id(), received, sent);
return 0;
@@ -122,7 +122,7 @@ worker(void *arg)
static int
consumer(void *arg)
{
- struct rte_event events[BATCH_SIZE] = {0};
+ struct rte_event events[BATCH_SIZE];
struct cons_data *data = (struct cons_data *)arg;
uint8_t event_dev_id = data->event_dev_id;
uint8_t event_port_id = data->event_port_id;
@@ -165,7 +165,7 @@ consumer(void *arg)
printf("deq_end = %"PRIu64", deq_start = %"PRIu64"\n",
deq_end, deq_start);
- printf("Consumer done! RX=%zu, time %"PRIu64"ms\n",
+ printf("Consumer done! RX=%"PRIu64", time %"PRIu64"ms\n",
num_packets,
(rte_get_timer_cycles() - start_time) / freq_khz);
done = 1;
@@ -188,7 +188,7 @@ producer(void *arg)
uint64_t enq_start, enq_end;
int k = 0;
struct rte_mbuf *m;
- struct rte_event producer_events[BATCH_SIZE] = {0};
+ struct rte_event producer_events[BATCH_SIZE];
struct rte_event *ev = &producer_events[0];
int l = 0;
struct rte_mbuf *mbufs[BATCH_SIZE];
@@ -263,7 +263,7 @@ producer(void *arg)
printf("Producer done. %"PRIu64" packets sent in %"PRIu64" cycles"
"(%f cycles/evt) (%f pkts/sec)\n",
- num_packets, enq_end-enq_start,
+ num_packets, enq_end - enq_start,
(float)(enq_end - enq_start)/(float)num_packets,
(float) (num_packets * rte_get_timer_hz()) /
(float) (enq_end - enq_start));
--
2.25.1
[PATCH v3] examples: add eventdev_producer_consumer example
The eventdev-producer-consumer application is a single-stage
producer-worker-consumer pipeline sample to mimic real-world applications.
It is useful in measuring performance impact when any eventdev
configuration is changed. Unlike test-eventdev, it allows configuring a
load balanced queue between the producer and workers and a single-link
queue between the workers and consumer. With test-eventdev, multiple worker
stages can be configured but there is no single consumer receiving events
from all the workers. It also does not require configuring TX/RX adapters
like in the case of eventdev_pipeline app.
Signed-off-by: Timothy McDaniel
---
V3: Fixed style and format issues, primarily those involving data types
whose size varies depending on whether we are building for 32 or
64 bit platforms.
V2: Disregard - forgot to resubmit entire patch
---
---
examples/eventdev_producer_consumer/Makefile | 22 +
examples/eventdev_producer_consumer/main.c| 670 ++
.../eventdev_producer_consumer/meson.build| 13 +
examples/meson.build | 1 +
4 files changed, 706 insertions(+)
create mode 100644 examples/eventdev_producer_consumer/Makefile
create mode 100644 examples/eventdev_producer_consumer/main.c
create mode 100644 examples/eventdev_producer_consumer/meson.build
diff --git a/examples/eventdev_producer_consumer/Makefile
b/examples/eventdev_producer_consumer/Makefile
new file mode 100644
index 00..761689eab7
--- /dev/null
+++ b/examples/eventdev_producer_consumer/Makefile
@@ -0,0 +1,22 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2016-2017 Intel Corporation.
+
+ifeq ($(RTE_SDK),)
+$(error "Please define RTE_SDK environment variable")
+endif
+
+# Default target, can be overridden by command line or environment
+RTE_TARGET ?= x86_64-native-linuxapp-gcc
+
+include $(RTE_SDK)/mk/rte.vars.mk
+
+# binary name
+APP = eventdev_producer_consumer
+
+# all source are stored in SRCS-y
+SRCS-y := main.c
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+include $(RTE_SDK)/mk/rte.extapp.mk
diff --git a/examples/eventdev_producer_consumer/main.c
b/examples/eventdev_producer_consumer/main.c
new file mode 100644
index 00..164bdf6f74
--- /dev/null
+++ b/examples/eventdev_producer_consumer/main.c
@@ -0,0 +1,670 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2022 Intel Corporation
+ */
+
+#include
+#include
+#include
+#include
+#include
+
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+
+#define BATCH_SIZE 32
+
+static unsigned int num_workers = 4;
+static bool g_is_mbuf;
+static uint64_t num_packets = (1L << 25); /* do ~32M packets */
+static int sched_type = RTE_SCHED_TYPE_ATOMIC;
+
+struct prod_data {
+ uint8_t event_dev_id;
+ uint8_t event_port_id;
+ int32_t qid;
+};
+
+struct cons_data {
+ uint8_t event_dev_id;
+ uint8_t event_port_id;
+};
+
+struct worker_data {
+ uint8_t event_dev_id;
+ int event_port_id;
+ int32_t qid;
+};
+
+static volatile int done;
+static int quiet;
+
+#define PORT_0 0
+#define QUEUE_0 0
+static struct rte_mempool *mp;
+
+static int
+worker(void *arg)
+{
+ struct rte_event rcv_events[BATCH_SIZE];
+
+ struct worker_data *data = (struct worker_data *)arg;
+ uint8_t event_dev_id = data->event_dev_id;
+ uint8_t event_port_id = data->event_port_id;
+ int32_t qid = data->qid;
+ uint64_t sent = 0, received = 0;
+ uint16_t n;
+
+ if (!quiet)
+ printf("Worker core %d started, portId=%d, sending to qid=%d\n",
+ rte_lcore_id(), event_port_id, qid);
+
+ while (!done) {
+ uint16_t k;
+ int npkts_to_send, npkts_sent = 0;
+ struct rte_event *ev;
+ uint64_t delay_start;
+
+ /* Cannot wait for IRQ here due to the way that
+* we check for when we are done.
+*/
+ n = rte_event_dequeue_burst(event_dev_id,
+ event_port_id,
+ rcv_events,
+ RTE_DIM(rcv_events),
+ 0);
+
+ if (n == 0) {
+ rte_pause();
+ continue;
+ } else if (!quiet)
+ printf("Worker received %d events(%"PRIu64" total)\n",
+ n, received);
+
+ delay_start = rte_rdtsc();
+ while (delay_start > rte_rdtsc())
+ ;
+
+ received += n;
+
+ ev = &rcv_events[0];
+ for (k = 0; k < n; k++) {
+ ev->queue_id = qid;
+ ev->op = RTE_EVENT_OP_FORWARD;
+ ev++;
+ }
+
+ ev = &rcv_events[0];
+ npkts_to_send = n, n
[PATCH v4] examples: add eventdev_producer_consumer example
The eventdev-producer-consumer application is a single-stage
producer-worker-consumer pipeline sample to mimic real-world applications.
It is useful in measuring performance impact when any eventdev
configuration is changed. Unlike test-eventdev, it allows configuring a
load balanced queue between the producer and workers and a single-link
queue between the workers and consumer. With test-eventdev, multiple worker
stages can be configured but there is no single consumer receiving events
from all the workers. It also does not require configuring TX/RX adapters
like in the case of eventdev_pipeline app.
Signed-off-by: Timothy McDaniel
---
V4: Fixed a coding style issue
V3: Fixed style and format issues, primarily those involving data types
whose size varies depending on whether we are building for 32 or
64 bit platforms.
V2: Disregard - forgot to resubmit entire patch
---
---
examples/eventdev_producer_consumer/Makefile | 22 +
examples/eventdev_producer_consumer/main.c| 671 ++
.../eventdev_producer_consumer/meson.build| 13 +
examples/meson.build | 1 +
4 files changed, 707 insertions(+)
create mode 100644 examples/eventdev_producer_consumer/Makefile
create mode 100644 examples/eventdev_producer_consumer/main.c
create mode 100644 examples/eventdev_producer_consumer/meson.build
diff --git a/examples/eventdev_producer_consumer/Makefile
b/examples/eventdev_producer_consumer/Makefile
new file mode 100644
index 00..761689eab7
--- /dev/null
+++ b/examples/eventdev_producer_consumer/Makefile
@@ -0,0 +1,22 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2016-2017 Intel Corporation.
+
+ifeq ($(RTE_SDK),)
+$(error "Please define RTE_SDK environment variable")
+endif
+
+# Default target, can be overridden by command line or environment
+RTE_TARGET ?= x86_64-native-linuxapp-gcc
+
+include $(RTE_SDK)/mk/rte.vars.mk
+
+# binary name
+APP = eventdev_producer_consumer
+
+# all source are stored in SRCS-y
+SRCS-y := main.c
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+include $(RTE_SDK)/mk/rte.extapp.mk
diff --git a/examples/eventdev_producer_consumer/main.c
b/examples/eventdev_producer_consumer/main.c
new file mode 100644
index 00..4c9f51d8c2
--- /dev/null
+++ b/examples/eventdev_producer_consumer/main.c
@@ -0,0 +1,671 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2022 Intel Corporation
+ */
+
+#include
+#include
+#include
+#include
+#include
+
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+
+#define BATCH_SIZE 32
+
+static unsigned int num_workers = 4;
+static bool g_is_mbuf;
+static uint64_t num_packets = (1L << 25); /* do ~32M packets */
+static int sched_type = RTE_SCHED_TYPE_ATOMIC;
+
+struct prod_data {
+ uint8_t event_dev_id;
+ uint8_t event_port_id;
+ int32_t qid;
+};
+
+struct cons_data {
+ uint8_t event_dev_id;
+ uint8_t event_port_id;
+};
+
+struct worker_data {
+ uint8_t event_dev_id;
+ int event_port_id;
+ int32_t qid;
+};
+
+static volatile int done;
+static int quiet;
+
+#define PORT_0 0
+#define QUEUE_0 0
+static struct rte_mempool *mp;
+
+static int
+worker(void *arg)
+{
+ struct rte_event rcv_events[BATCH_SIZE];
+
+ struct worker_data *data = (struct worker_data *)arg;
+ uint8_t event_dev_id = data->event_dev_id;
+ uint8_t event_port_id = data->event_port_id;
+ int32_t qid = data->qid;
+ uint64_t sent = 0, received = 0;
+ uint16_t n;
+
+ if (!quiet)
+ printf("Worker core %d started, portId=%d, sending to qid=%d\n",
+ rte_lcore_id(), event_port_id, qid);
+
+ while (!done) {
+ uint16_t k;
+ int npkts_to_send, npkts_sent = 0;
+ struct rte_event *ev;
+ uint64_t delay_start;
+
+ /* Cannot wait for IRQ here due to the way that
+* we check for when we are done.
+*/
+ n = rte_event_dequeue_burst(event_dev_id,
+ event_port_id,
+ rcv_events,
+ RTE_DIM(rcv_events),
+ 0);
+
+ if (n == 0) {
+ rte_pause();
+ continue;
+ } else if (!quiet)
+ printf("Worker received %d events(%"PRIu64" total)\n",
+ n, received);
+
+ delay_start = rte_rdtsc();
+ while (delay_start > rte_rdtsc())
+ ;
+
+ received += n;
+
+ ev = &rcv_events[0];
+ for (k = 0; k < n; k++) {
+ ev->queue_id = qid;
+ ev->op = RTE_EVENT_OP_FORWARD;
+ ev++;
+ }
+
+ ev = &rcv_events[0];
+
RE: [PATCH v2] net/ice: support disabling ACL engine in DCF via devargs
> -Original Message- > From: Zeng, ZhichaoX > Sent: Wednesday, August 17, 2022 4:21 PM > To: dev@dpdk.org > Cc: Yang, Qiming ; Zeng, ZhichaoX > ; Zhang, Qi Z > Subject: [PATCH v2] net/ice: support disabling ACL engine in DCF via devargs > > From: Zhichao Zeng > > Support disabling DCF ACL engine via devarg "acl=off" in cmdline, aiming to > shorten the DCF startup time. > > Signed-off-by: Zhichao Zeng Acked-by: Qi Zhang Applied to dpdk-next-net-intel. Thanks Qi
RE: [PATCH v2 1/3] ethdev: introduce pool sort capability
Hi Hanumanth,
> -Original Message-
> From: Hanumanth Pothula
> Sent: Saturday, August 13, 2022 1:25 AM
> To: Thomas Monjalon ; Ferruh Yigit
> ; Andrew Rybchenko
>
> Cc: dev@dpdk.org; Ding, Xuan ; Wu, WenxuanX
> ; Li, Xiaoyun ;
> step...@networkplumber.org; Wang, YuanX ;
> m...@ashroe.eu; Zhang, Yuying ; Zhang, Qi Z
> ; viachesl...@nvidia.com; jer...@marvell.com;
> ndabilpu...@marvell.com; Hanumanth Pothula
> Subject: [PATCH v2 1/3] ethdev: introduce pool sort capability
>
> Presently, the 'Buffer Split' feature supports sending multiple segments of
> the received packet to PMD, which programs the HW to receive the packet in
> segments from different pools.
>
> This patch extends the feature to support the pool sort capability.
> Some of the HW has support for choosing memory pools based on the
> packet's size. The pool sort capability allows PMD to choose a memory pool
> based on the packet's length.
>
> This is often useful for saving the memory where the application can create a
> different pool to steer the specific size of the packet, thus enabling
> effective
> use of memory.
>
> For example, let's say HW has a capability of three pools,
> - pool-1 size is 2K
> - pool-2 size is > 2K and < 4K
> - pool-3 size is > 4K
> Here,
> pool-1 can accommodate packets with sizes < 2K
> pool-2 can accommodate packets with sizes > 2K and < 4K
> pool-3 can accommodate packets with sizes > 4K
>
> With pool sort capability enabled in SW, an application may create three
> pools of different sizes and send them to PMD. Allowing PMD to program
> HW based on packet lengths. So that packets with less than 2K are received
> on pool-1, packets with lengths between 2K and 4K are received on pool-2
> and finally packets greater than 4K are received on pool-3.
>
> The following two capabilities are added to the rte_eth_rxseg_capa structure,
> 1. pool_sort --> tells pool sort capability is supported by HW.
> 2. max_npool --> max number of pools supported by HW.
>
> Defined new structure rte_eth_rxseg_sort, to be used only when pool sort
> capability is present. If required this may be extended further to support
> more configurations.
>
> Signed-off-by: Hanumanth Pothula
>
> v2:
> - Along with spec changes, uploading testpmd and driver changes.
Thanks for CCing. It's an interesting feature.
But I have one question here:
Buffer split is for split receiving packets into multiple segments, while pool
sort supports
PMD to put the receiving packets into different pools according to packet size.
Every packet is still intact.
So, at this level, pool sort does not belong to buffer split.
And you already use a different function to check pool sort rather than check
buffer split.
Should a new RX offload be introduced? like "RTE_ETH_RX_OFFLOAD_POOL_SORT".
> ---
> lib/ethdev/rte_ethdev.c | 87 +++--
> lib/ethdev/rte_ethdev.h | 45 +++--
> 2 files changed, 118 insertions(+), 14 deletions(-)
>
> diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c index
> 1979dc0850..7fd5443eb8 100644
> --- a/lib/ethdev/rte_ethdev.c
> +++ b/lib/ethdev/rte_ethdev.c
> @@ -1635,7 +1635,55 @@ rte_eth_dev_is_removed(uint16_t port_id) }
>
> static int
> -rte_eth_rx_queue_check_split(const struct rte_eth_rxseg_split *rx_seg,
> +rte_eth_rx_queue_check_sort(const struct rte_eth_rxseg *rx_seg,
> + uint16_t n_seg, uint32_t *mbp_buf_size,
> + const struct rte_eth_dev_info *dev_info) {
> + const struct rte_eth_rxseg_capa *seg_capa = &dev_info-
> >rx_seg_capa;
> + uint16_t seg_idx;
> +
> + if (!seg_capa->multi_pools || n_seg > seg_capa->max_npool) {
> + RTE_ETHDEV_LOG(ERR,
> +"Invalid capabilities, multi_pools:%d differnt
> length segments %u exceed supported %u\n",
> +seg_capa->multi_pools, n_seg, seg_capa-
> >max_nseg);
> + return -EINVAL;
> + }
> +
> + for (seg_idx = 0; seg_idx < n_seg; seg_idx++) {
> + struct rte_mempool *mpl = rx_seg[seg_idx].sort.mp;
> + uint32_t length = rx_seg[seg_idx].sort.length;
> +
> + if (mpl == NULL) {
> + RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
> + return -EINVAL;
> + }
> +
> + if (mpl->private_data_size <
> + sizeof(struct rte_pktmbuf_pool_private)) {
> + RTE_ETHDEV_LOG(ERR,
> +"%s private_data_size %u < %u\n",
> +mpl->name, mpl->private_data_size,
> +(unsigned int)sizeof
> + (struct rte_pktmbuf_pool_private));
> + return -ENOSPC;
> + }
> +
> + *mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
> + length = length != 0 ? length : (*m
[PATCH 0/8] add multi queue support to vDPA ifc driver
Add multi queue support to vDPA ifc driver. Multi queue support for virtio-net device and virtio-blk device. Andy Pei (7): vdpa/ifc: add multi queue suppoort vdpa/ifc: set max queues according to HW spec vdpa/ifc: write queue count to MQ register vdpa/ifc: only configure enabled queue vdpa/ifc: set vring state callback update data path vhost: configure device when any queue is ready for BLK device vhost: vDPA BLK devices configure device when all queue callfds are set Huang Wei (1): vdpa/ifc: add new device ID drivers/vdpa/ifc/base/ifcvf.c | 24 - drivers/vdpa/ifc/base/ifcvf.h | 8 -- drivers/vdpa/ifc/ifcvf_vdpa.c | 59 ++-- lib/vhost/vhost_user.c| 63 ++- 4 files changed, 119 insertions(+), 35 deletions(-) -- 1.8.3.1
[PATCH 1/8] vdpa/ifc: add new device ID
From: Huang Wei
Add new device id to support IFCVF_NET_TRANSITIONAL_DEVICE_ID (0x1000).
Signed-off-by: Huang Wei
Signed-off-by: Andy Pei
---
drivers/vdpa/ifc/base/ifcvf.h | 4 +++-
drivers/vdpa/ifc/ifcvf_vdpa.c | 9 -
2 files changed, 11 insertions(+), 2 deletions(-)
diff --git a/drivers/vdpa/ifc/base/ifcvf.h b/drivers/vdpa/ifc/base/ifcvf.h
index 9d95aac..7ede738 100644
--- a/drivers/vdpa/ifc/base/ifcvf.h
+++ b/drivers/vdpa/ifc/base/ifcvf.h
@@ -12,11 +12,13 @@
#define IFCVF_BLK 1
#define IFCVF_VENDOR_ID 0x1AF4
-#define IFCVF_NET_DEVICE_ID 0x1041
+#define IFCVF_NET_MODERN_DEVICE_ID 0x1041
#define IFCVF_BLK_MODERN_DEVICE_ID 0x1042
+#define IFCVF_NET_TRANSITIONAL_DEVICE_ID0x1000
#define IFCVF_BLK_TRANSITIONAL_DEVICE_ID0x1001
#define IFCVF_SUBSYS_VENDOR_ID 0x8086
#define IFCVF_SUBSYS_DEVICE_ID 0x001A
+#define IFCVF_NET_DEVICE_ID 0x0001
#define IFCVF_BLK_DEVICE_ID 0x0002
#define IFCVF_MAX_QUEUES 1
diff --git a/drivers/vdpa/ifc/ifcvf_vdpa.c b/drivers/vdpa/ifc/ifcvf_vdpa.c
index ac42de9..61d0250 100644
--- a/drivers/vdpa/ifc/ifcvf_vdpa.c
+++ b/drivers/vdpa/ifc/ifcvf_vdpa.c
@@ -1684,13 +1684,20 @@ struct rte_vdpa_dev_info dev_info[] = {
static const struct rte_pci_id pci_id_ifcvf_map[] = {
{ .class_id = RTE_CLASS_ANY_ID,
.vendor_id = IFCVF_VENDOR_ID,
- .device_id = IFCVF_NET_DEVICE_ID,
+ .device_id = IFCVF_NET_MODERN_DEVICE_ID,
.subsystem_vendor_id = IFCVF_SUBSYS_VENDOR_ID,
.subsystem_device_id = IFCVF_SUBSYS_DEVICE_ID,
},
{ .class_id = RTE_CLASS_ANY_ID,
.vendor_id = IFCVF_VENDOR_ID,
+ .device_id = IFCVF_NET_TRANSITIONAL_DEVICE_ID,
+ .subsystem_vendor_id = IFCVF_SUBSYS_VENDOR_ID,
+ .subsystem_device_id = IFCVF_NET_DEVICE_ID,
+ },
+
+ { .class_id = RTE_CLASS_ANY_ID,
+ .vendor_id = IFCVF_VENDOR_ID,
.device_id = IFCVF_BLK_TRANSITIONAL_DEVICE_ID,
.subsystem_vendor_id = IFCVF_SUBSYS_VENDOR_ID,
.subsystem_device_id = IFCVF_BLK_DEVICE_ID,
--
1.8.3.1
[PATCH 2/8] vdpa/ifc: add multi queue suppoort
Enable VHOST_USER_PROTOCOL_F_MQ feature.
ExposeIFCVF_MQ_OFFSET register to enable multi queue.
Signed-off-by: Andy Pei
Signed-off-by: Huang Wei
---
drivers/vdpa/ifc/base/ifcvf.c | 5 +
drivers/vdpa/ifc/base/ifcvf.h | 2 ++
drivers/vdpa/ifc/ifcvf_vdpa.c | 1 +
3 files changed, 8 insertions(+)
diff --git a/drivers/vdpa/ifc/base/ifcvf.c b/drivers/vdpa/ifc/base/ifcvf.c
index f1e1474..34c8226 100644
--- a/drivers/vdpa/ifc/base/ifcvf.c
+++ b/drivers/vdpa/ifc/base/ifcvf.c
@@ -90,6 +90,11 @@
if (!hw->lm_cfg)
WARNINGOUT("HW support live migration not support!\n");
+ if (hw->mem_resource[4].addr)
+ hw->mq_cfg = hw->mem_resource[4].addr + IFCVF_MQ_OFFSET;
+ else
+ hw->mq_cfg = NULL;
+
if (hw->common_cfg == NULL || hw->notify_base == NULL ||
hw->isr == NULL || hw->dev_cfg == NULL) {
DEBUGOUT("capability incomplete\n");
diff --git a/drivers/vdpa/ifc/base/ifcvf.h b/drivers/vdpa/ifc/base/ifcvf.h
index 7ede738..ad505f1 100644
--- a/drivers/vdpa/ifc/base/ifcvf.h
+++ b/drivers/vdpa/ifc/base/ifcvf.h
@@ -50,6 +50,7 @@
#define IFCVF_LM_CFG_SIZE 0x40
#define IFCVF_LM_RING_STATE_OFFSET 0x20
+#define IFCVF_MQ_OFFSET0x28
#define IFCVF_LM_LOGGING_CTRL 0x0
@@ -149,6 +150,7 @@ struct ifcvf_hw {
u16*notify_base;
u16*notify_addr[IFCVF_MAX_QUEUES * 2];
u8 *lm_cfg;
+ u8 *mq_cfg;
struct vring_info vring[IFCVF_MAX_QUEUES * 2];
u8 nr_vring;
int device_type;
diff --git a/drivers/vdpa/ifc/ifcvf_vdpa.c b/drivers/vdpa/ifc/ifcvf_vdpa.c
index 61d0250..2d165c0 100644
--- a/drivers/vdpa/ifc/ifcvf_vdpa.c
+++ b/drivers/vdpa/ifc/ifcvf_vdpa.c
@@ -1248,6 +1248,7 @@ struct rte_vdpa_dev_info {
1ULL << VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD | \
1ULL << VHOST_USER_PROTOCOL_F_HOST_NOTIFIER | \
1ULL << VHOST_USER_PROTOCOL_F_LOG_SHMFD | \
+1ULL << VHOST_USER_PROTOCOL_F_MQ | \
1ULL << VHOST_USER_PROTOCOL_F_STATUS)
#define VDPA_BLK_PROTOCOL_FEATURES \
--
1.8.3.1
[PATCH 3/8] vdpa/ifc: set max queues according to HW spec
Set max_queues according to virtio HW spec.
For virtio BLK device, set max_queues to the value of "num_queues".
"num_queues" is element of struct virtio_blk_config.
Signed-off-by: Andy Pei
Signed-off-by: Huang Wei
---
drivers/vdpa/ifc/base/ifcvf.h | 2 +-
drivers/vdpa/ifc/ifcvf_vdpa.c | 18 +-
2 files changed, 18 insertions(+), 2 deletions(-)
diff --git a/drivers/vdpa/ifc/base/ifcvf.h b/drivers/vdpa/ifc/base/ifcvf.h
index ad505f1..c17bf2a 100644
--- a/drivers/vdpa/ifc/base/ifcvf.h
+++ b/drivers/vdpa/ifc/base/ifcvf.h
@@ -21,7 +21,7 @@
#define IFCVF_NET_DEVICE_ID 0x0001
#define IFCVF_BLK_DEVICE_ID 0x0002
-#define IFCVF_MAX_QUEUES 1
+#define IFCVF_MAX_QUEUES 32
#ifndef VIRTIO_F_IOMMU_PLATFORM
#define VIRTIO_F_IOMMU_PLATFORM33
diff --git a/drivers/vdpa/ifc/ifcvf_vdpa.c b/drivers/vdpa/ifc/ifcvf_vdpa.c
index 2d165c0..34aea6c 100644
--- a/drivers/vdpa/ifc/ifcvf_vdpa.c
+++ b/drivers/vdpa/ifc/ifcvf_vdpa.c
@@ -26,6 +26,18 @@
#include "base/ifcvf.h"
+/**
+** RTE_MAX() and RTE_MIN() cannot be used since braced-group within
+** expression allowed only inside a function, but MAX() is used as
+** a number of elements in array.
+**/
+#ifndef MAX
+#define MAX(v1, v2)((v1) > (v2) ? (v1) : (v2))
+#endif
+#ifndef MIN
+#define MIN(v1, v2)((v1) < (v2) ? (v1) : (v2))
+#endif
+
RTE_LOG_REGISTER(ifcvf_vdpa_logtype, pmd.vdpa.ifcvf, NOTICE);
#define DRV_LOG(level, fmt, args...) \
rte_log(RTE_LOG_ ## level, ifcvf_vdpa_logtype, \
@@ -1559,7 +1571,6 @@ struct rte_vdpa_dev_info dev_info[] = {
}
internal->configured = 0;
- internal->max_queues = IFCVF_MAX_QUEUES;
features = ifcvf_get_features(&internal->hw);
device_id = ifcvf_pci_get_device_type(pci_dev);
@@ -1570,6 +1581,8 @@ struct rte_vdpa_dev_info dev_info[] = {
if (device_id == VIRTIO_ID_NET) {
internal->hw.device_type = IFCVF_NET;
+ internal->max_queues = MIN(IFCVF_MAX_QUEUES,
+ (internal->hw.common_cfg->num_queues - 1)/2);
internal->features = features &
~(1ULL << VIRTIO_F_IOMMU_PLATFORM);
internal->features |= dev_info[IFCVF_NET].features;
@@ -1609,6 +1622,9 @@ struct rte_vdpa_dev_info dev_info[] = {
internal->hw.blk_cfg->geometry.sectors);
DRV_LOG(DEBUG, "num_queues: 0x%08x",
internal->hw.blk_cfg->num_queues);
+
+ internal->max_queues = MIN(IFCVF_MAX_QUEUES,
+ internal->hw.blk_cfg->num_queues);
}
list->internal = internal;
--
1.8.3.1
[PATCH 4/8] vdpa/ifc: write queue count to MQ register
Write queue count to IFCVF_MQ_OFFSET register
to enable multi queue feature.
Signed-off-by: Andy Pei
Signed-off-by: Huang Wei
---
drivers/vdpa/ifc/base/ifcvf.c | 14 ++
1 file changed, 14 insertions(+)
diff --git a/drivers/vdpa/ifc/base/ifcvf.c b/drivers/vdpa/ifc/base/ifcvf.c
index 34c8226..1b50df6 100644
--- a/drivers/vdpa/ifc/base/ifcvf.c
+++ b/drivers/vdpa/ifc/base/ifcvf.c
@@ -198,6 +198,19 @@
IFCVF_WRITE_REG32(val >> 32, hi);
}
+STATIC void
+ifcvf_enable_multiqueue(struct ifcvf_hw *hw, u16 nr_queue_pair)
+{
+ u8 *mq_cfg;
+
+ if (hw->device_type == IFCVF_NET)
+ nr_queue_pair = (nr_queue_pair + 1) / 2;
+
+ mq_cfg = hw->mq_cfg;
+ if (mq_cfg)
+ *(u32 *)mq_cfg = nr_queue_pair;
+}
+
STATIC int
ifcvf_hw_enable(struct ifcvf_hw *hw)
{
@@ -215,6 +228,7 @@
return -1;
}
+ ifcvf_enable_multiqueue(hw, hw->nr_vring);
for (i = 0; i < hw->nr_vring; i++) {
IFCVF_WRITE_REG16(i, &cfg->queue_select);
io_write64_twopart(hw->vring[i].desc, &cfg->queue_desc_lo,
--
1.8.3.1
[PATCH 5/8] vdpa/ifc: only configure enabled queue
when configure the hardware queue, we only configure queues which
have been enabled by vhost.
Signed-off-by: Andy Pei
---
drivers/vdpa/ifc/base/ifcvf.c | 5 -
drivers/vdpa/ifc/ifcvf_vdpa.c | 4
2 files changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/vdpa/ifc/base/ifcvf.c b/drivers/vdpa/ifc/base/ifcvf.c
index 1b50df6..ca5f677 100644
--- a/drivers/vdpa/ifc/base/ifcvf.c
+++ b/drivers/vdpa/ifc/base/ifcvf.c
@@ -230,6 +230,8 @@
ifcvf_enable_multiqueue(hw, hw->nr_vring);
for (i = 0; i < hw->nr_vring; i++) {
+ if (!hw->vring[i].enable)
+ continue;
IFCVF_WRITE_REG16(i, &cfg->queue_select);
io_write64_twopart(hw->vring[i].desc, &cfg->queue_desc_lo,
&cfg->queue_desc_hi);
@@ -264,7 +266,8 @@
notify_off = IFCVF_READ_REG16(&cfg->queue_notify_off);
hw->notify_addr[i] = (void *)((u8 *)hw->notify_base +
notify_off * hw->notify_off_multiplier);
- IFCVF_WRITE_REG16(1, &cfg->queue_enable);
+ if (hw->vring[i].enable)
+ IFCVF_WRITE_REG16(1, &cfg->queue_enable);
}
return 0;
diff --git a/drivers/vdpa/ifc/ifcvf_vdpa.c b/drivers/vdpa/ifc/ifcvf_vdpa.c
index 34aea6c..a62bcec 100644
--- a/drivers/vdpa/ifc/ifcvf_vdpa.c
+++ b/drivers/vdpa/ifc/ifcvf_vdpa.c
@@ -290,6 +290,8 @@ struct rte_vdpa_dev_info {
rte_vhost_get_negotiated_features(vid, &hw->req_features);
for (i = 0; i < nr_vring; i++) {
+ if (!hw->vring[i].enable)
+ continue;
rte_vhost_get_vhost_vring(vid, i, &vq);
gpa = hva_to_gpa(vid, (uint64_t)(uintptr_t)vq.desc);
if (gpa == 0) {
@@ -505,6 +507,8 @@ struct rte_vdpa_dev_info {
vring.kickfd = -1;
for (qid = 0; qid < q_num; qid++) {
+ if (!hw->vring[qid].enable)
+ continue;
ev.events = EPOLLIN | EPOLLPRI;
rte_vhost_get_vhost_vring(internal->vid, qid, &vring);
ev.data.u64 = qid | (uint64_t)vring.kickfd << 32;
--
1.8.3.1
[PATCH 6/8] vdpa/ifc: set vring state callback update data path
To support multi queue, in the case that first queue is ready
and device is configured, when more queues need to be configured,
we just close and restart data path.
This also fix the situation that using set_vring_state callback
to disable one queue will cause all vfio interrupts being disabled.
Signed-off-by: Andy Pei
---
drivers/vdpa/ifc/ifcvf_vdpa.c | 27 +--
1 file changed, 9 insertions(+), 18 deletions(-)
diff --git a/drivers/vdpa/ifc/ifcvf_vdpa.c b/drivers/vdpa/ifc/ifcvf_vdpa.c
index a62bcec..94c8ef1 100644
--- a/drivers/vdpa/ifc/ifcvf_vdpa.c
+++ b/drivers/vdpa/ifc/ifcvf_vdpa.c
@@ -1286,8 +1286,6 @@ struct rte_vdpa_dev_info {
struct internal_list *list;
struct ifcvf_internal *internal;
struct ifcvf_hw *hw;
- struct ifcvf_pci_common_cfg *cfg;
- int ret = 0;
vdev = rte_vhost_get_vdpa_device(vid);
list = find_internal_resource_by_vdev(vdev);
@@ -1303,27 +1301,20 @@ struct rte_vdpa_dev_info {
}
hw = &internal->hw;
- if (!internal->configured)
- goto exit;
- cfg = hw->common_cfg;
- IFCVF_WRITE_REG16(vring, &cfg->queue_select);
- IFCVF_WRITE_REG16(!!state, &cfg->queue_enable);
+ hw->vring[vring].enable = !!state;
- if (!state && hw->vring[vring].enable) {
- ret = vdpa_disable_vfio_intr(internal);
- if (ret)
- return ret;
- }
+ if (!internal->configured)
+ goto exit;
- if (state && !hw->vring[vring].enable) {
- ret = vdpa_enable_vfio_intr(internal, false);
- if (ret)
- return ret;
- }
+ /* close data path */
+ rte_atomic32_set(&internal->dev_attached, 0);
+ update_datapath(internal);
+ /* restart data path */
+ rte_atomic32_set(&internal->dev_attached, 1);
+ update_datapath(internal);
exit:
- hw->vring[vring].enable = !!state;
return 0;
}
--
1.8.3.1
[PATCH 7/8] vhost: configure device when any queue is ready for BLK device
When boot from virtio blk device, seabois in QEMU only enables one queue.
To work in this scenario, vDPA BLK device back-end conf_dev when any
queue is ready.
Signed-off-by: Andy Pei
---
lib/vhost/vhost_user.c | 56 +++---
1 file changed, 44 insertions(+), 12 deletions(-)
diff --git a/lib/vhost/vhost_user.c b/lib/vhost/vhost_user.c
index 4ad28ba..b65fba3 100644
--- a/lib/vhost/vhost_user.c
+++ b/lib/vhost/vhost_user.c
@@ -1451,6 +1451,25 @@
#define VIRTIO_BUILTIN_NUM_VQS_TO_BE_READY 2u
static int
+virtio_has_queue_ready(struct virtio_net *dev)
+{
+ struct vhost_virtqueue *vq;
+ uint32_t i, nr_vring = dev->nr_vring;
+
+ if (!dev->nr_vring)
+ return 0;
+
+ for (i = 0; i < nr_vring; i++) {
+ vq = dev->virtqueue[i];
+
+ if (vq_is_ready(dev, vq))
+ return 1;
+ }
+
+ return 0;
+}
+
+static int
virtio_is_ready(struct virtio_net *dev)
{
struct vhost_virtqueue *vq;
@@ -3167,9 +3186,33 @@ static int is_vring_iotlb(struct virtio_net *dev,
if (unlock_required)
vhost_user_unlock_all_queue_pairs(dev);
- if (ret != 0 || !virtio_is_ready(dev))
+ if (ret != 0)
goto out;
+ vdpa_dev = dev->vdpa_dev;
+ if (vdpa_dev) {
+ if (vdpa_dev->ops->get_dev_type) {
+ ret = vdpa_dev->ops->get_dev_type(vdpa_dev, &vdpa_type);
+ if (ret) {
+ VHOST_LOG_CONFIG(dev->ifname, ERR,
+ "failed to get vdpa dev type.\n");
+ ret = -1;
+ goto out;
+ }
+ } else {
+ vdpa_type = RTE_VHOST_VDPA_DEVICE_TYPE_NET;
+ }
+ }
+
+ if (!virtio_is_ready(dev)) {
+ if (vdpa_type == RTE_VHOST_VDPA_DEVICE_TYPE_BLK) {
+ if (!virtio_has_queue_ready(dev))
+ goto out;
+ } else {
+ goto out;
+ }
+ }
+
/*
* Virtio is now ready. If not done already, it is time
* to notify the application it can process the rings and
@@ -3181,20 +3224,9 @@ static int is_vring_iotlb(struct virtio_net *dev,
dev->flags |= VIRTIO_DEV_RUNNING;
}
- vdpa_dev = dev->vdpa_dev;
if (!vdpa_dev)
goto out;
- if (vdpa_dev->ops->get_dev_type) {
- ret = vdpa_dev->ops->get_dev_type(vdpa_dev, &vdpa_type);
- if (ret) {
- VHOST_LOG_CONFIG(dev->ifname, ERR, "failed to get vdpa
dev type.\n");
- ret = -1;
- goto out;
- }
- } else {
- vdpa_type = RTE_VHOST_VDPA_DEVICE_TYPE_NET;
- }
if (vdpa_type == RTE_VHOST_VDPA_DEVICE_TYPE_BLK
&& request != VHOST_USER_SET_VRING_CALL)
goto out;
--
1.8.3.1
[PATCH 8/8] vhost: vDPA BLK devices configure device when all queue callfds are set
In the virtio blk vDPA live migration use case, before the live
migration process, QEMU will set call fd to vDPA back-end. QEMU
and vDPA back-end stand by until live migration starts.
During live migration process, QEMU sets kick fd and a new call
fd. However, after the kick fd is set to the vDPA back-end, the
vDPA back-end configures device and data path starts. The new
call fd will cause some kind of "re-configuration", this kind
of "re-configuration" cause IO drop.
After this patch, vDPA back-end configures device after kick fd
and call fd are well set and make sure no IO drops.
This patch only impact virtio blk vDPA device and does not impact
net device.
Signed-off-by: Andy Pei
---
lib/vhost/vhost_user.c | 13 ++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/lib/vhost/vhost_user.c b/lib/vhost/vhost_user.c
index b65fba3..568030a 100644
--- a/lib/vhost/vhost_user.c
+++ b/lib/vhost/vhost_user.c
@@ -2994,6 +2994,7 @@ static int is_vring_iotlb(struct virtio_net *dev,
uint32_t vdpa_type = 0;
uint32_t request;
uint32_t i;
+ uint16_t blk_call_fd;
dev = get_device(vid);
if (dev == NULL)
@@ -3227,9 +3228,15 @@ static int is_vring_iotlb(struct virtio_net *dev,
if (!vdpa_dev)
goto out;
- if (vdpa_type == RTE_VHOST_VDPA_DEVICE_TYPE_BLK
- && request != VHOST_USER_SET_VRING_CALL)
- goto out;
+ if (vdpa_type == RTE_VHOST_VDPA_DEVICE_TYPE_BLK) {
+ if (request == VHOST_USER_SET_VRING_CALL) {
+ blk_call_fd = ctx.msg.payload.u64 &
VHOST_USER_VRING_IDX_MASK;
+ if (blk_call_fd != dev->nr_vring - 1)
+ goto out;
+ } else {
+ goto out;
+ }
+ }
if (!(dev->flags & VIRTIO_DEV_VDPA_CONFIGURED)) {
if (vdpa_dev->ops->dev_conf(dev->vid))
--
1.8.3.1
RE: [EXT] [PATCH v3 0/3] net/octeon_ep: rename driver and add features
> -Original Message-
> From: Sathesh Edara
> Sent: Monday, August 22, 2022 2:10 AM
> To: Satananda Burla ; Jerin Jacob Kollanukkaran
> ; Sathesh B Edara
> Cc: dev@dpdk.org
> Subject: [EXT] [PATCH v3 0/3] net/octeon_ep: rename driver and add
> features
>
> External Email
>
> --
> This patch set renames the net/octeontx_ep driver to net/octeon_ep and
> 2nd and 3rd patches add support for basic stats and link status.
>
> Changes in v3:
> - Updated commit messaage.
> - Updated deprecation.rst.
>
> Changes in v2:
> Added new features in the renamed driver.
>
> Sathesh Edara (3):
> net/octeontx_ep: rename as octeon_ep
> net/octeon_ep: support basic stats
> net/octeon_ep: support link status
>
> MAINTAINERS | 6 +-
> .../{octeontx_ep.ini => octeon_ep.ini}| 4 +-
> doc/guides/nics/index.rst | 2 +-
> .../nics/{octeontx_ep.rst => octeon_ep.rst} | 4 +-
> doc/guides/rel_notes/deprecation.rst | 5 --
> drivers/net/meson.build | 2 +-
> .../{octeontx_ep => octeon_ep}/meson.build| 0
> .../{octeontx_ep => octeon_ep}/otx2_ep_vf.c | 0
> .../{octeontx_ep => octeon_ep}/otx2_ep_vf.h | 0
> .../otx_ep_common.h | 0
> .../otx_ep_ethdev.c | 69 +++
> .../{octeontx_ep => octeon_ep}/otx_ep_rxtx.c | 0 .../{octeontx_ep =>
> octeon_ep}/otx_ep_rxtx.h | 0
> .../{octeontx_ep => octeon_ep}/otx_ep_vf.c| 0
> .../{octeontx_ep => octeon_ep}/otx_ep_vf.h| 0
> .../{octeontx_ep => octeon_ep}/version.map| 0
> 16 files changed, 79 insertions(+), 13 deletions(-) rename
> doc/guides/nics/features/{octeontx_ep.ini => octeon_ep.ini} (64%) rename
> doc/guides/nics/{octeontx_ep.rst => octeon_ep.rst} (87%) rename
> drivers/net/{octeontx_ep => octeon_ep}/meson.build (100%) rename
> drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.c (100%) rename
> drivers/net/{octeontx_ep => octeon_ep}/otx2_ep_vf.h (100%) rename
> drivers/net/{octeontx_ep => octeon_ep}/otx_ep_common.h (100%)
> rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_ethdev.c (86%)
> rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.c (100%)
> rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_rxtx.h (100%)
> rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.c (100%)
> rename drivers/net/{octeontx_ep => octeon_ep}/otx_ep_vf.h (100%)
> rename drivers/net/{octeontx_ep => octeon_ep}/version.map (100%)
>
> --
> 2.36.1
Series-acked-by: Veerasenareddy Burru
<>
RE: [EXT] Re: [PATCH] crypto/ccp: Check for the NULL pointer after calling rte_malloc
[Public]
+ sunil
-Original Message-
From: Akhil Goyal
Sent: Tuesday, August 16, 2022 9:31 PM
To: Namburu, Chandu-babu ; 835703...@qq.com
Cc: dev@dpdk.org; David Marchand
Subject: RE: [EXT] Re: [PATCH] crypto/ccp: Check for the NULL pointer after
calling rte_malloc
Hi,
Could you please reply to David and Stephen's comments?
Regards,
Akhil
> On Wed, Jul 20, 2022 at 8:29 AM Namburu, Chandu-babu
> wrote:
> > From: Shiqi Liu <835703...@qq.com>
> >
> > As the possible failure of the rte_malloc(), the not_checked and
> > checked could
> be NULL pointer.
> > Therefore, it should be better to check it in order to avoid the
> > dereference of
> the NULL pointer.
> >
> > Fixes: 09a0fd736a0 ("crypto/ccp: enable IOMMU")
> > Signed-off-by: Shiqi Liu <835703...@qq.com>
>
> This sha_ctx variable and its accesses are suspicious.
>
> It seems to be used as some kind of intermediate buffer, but I fail to
> see the need.
> Can't the existing code rely on sess->auth.ctx ?
>
> There is also a suspicious mention (in ccp_perform_sha) of sha_ctx but
> with no calling rte_mem_virt2iova().
>
>
> --
> David Marchand
