Re: [PATCH] devtool: fix falsely reporting from checkpatch

[Andre Muezerie]

On Tue, Jan 28, 2025 at 04:02:27PM +0100, David Marchand wrote:
> Hi,
> 
> On Mon, Jan 20, 2025 at 12:27 PM WanRenyong  wrote:
> >
> > When executes the check_packed_attributes function in checkpatch,
> > if __rte_packed_begin or __rte_packed_end appear in the context
> > of a patch file, there may be a situation where the counts of
> > __rte_packed_begin and __rte_packed_end do not match, causing
> > checkpatch to return a failure.
> > This patch fixes this issue by only counting the lines in the
> > patch file that start with a + and include either
> > __rte_packed_begin or __rte_packed_end.
> >
> > Signed-off-by: WanRenyong 
> 
> Thanks for proposing this fix.
> 
> When sending such fixes, don't forget to add a Fixes: and Cc:
> author/maintainers by using --cc-cmd devtools/get-maintainers.sh
> 
> 
> Adding Akhil, André and Thomas in the loop.
> 
> I also had some concern about the check:
> https://inbox.dpdk.org/dev/CAJFAV8w=s1L-WYk+Qv-B+Mn6eAwKrB=GTz6hU--ZoLrJsz7=d...@mail.gmail.com/
> 
> But I merged the check untouched as nobody else seemed to object.
> Checkpatch warnings are known to have false positive and such false
> positives are "filtered" by subtree maintainers.
> 
> 
> > ---
> >  devtools/checkpatches.sh | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/devtools/checkpatches.sh b/devtools/checkpatches.sh
> > index 003bb49e04..2e228b7f92 100755
> > --- a/devtools/checkpatches.sh
> > +++ b/devtools/checkpatches.sh
> > @@ -384,8 +384,8 @@ check_packed_attributes() { # 
> > res=1
> > fi
> >
> > -   begin_count=$(grep '__rte_packed_begin' "$1" | wc -l)
> > -   end_count=$(grep '__rte_packed_end' "$1" | wc -l)
> > +   begin_count=$(grep -E '^\+.*__rte_packed_begin' "$1" | wc -l)
> > +   end_count=$(grep -E '^\+.*__rte_packed_end' "$1" | wc -l)
> > if [ $begin_count != $end_count ]; then
> > echo "__rte_packed_begin and __rte_packed_end should always 
> > be used in pairs."
> > res=1
> 
> I don't think this change is any better.
> 
> There is a good chance that a patch touching just a first line of a
> structure definition won't come along a line touching the last line of
> the struct.
> My suggestion (if we want to avoid those non useful warning) would be
> to just remove the counting stuff in the check.
> 
> Opinions?
> 
> 
> -- 
> David Marchand

Yes, it was known that false positives could happen, as is the case with many 
other checks used today. While I was working on the patchset that added this 
new check it helped me identify some issues with my changes. I feel I benefited 
from this check at that time, but if people feel that it is causing more harm 
than good, I am not opposed to having it removed.

Re: [PATCH v1 00/31] net/ntnic: bugfixes and refactoring

[Stephen Hemminger]

On Tue, 21 Jan 2025 18:07:38 +0100
Serhii Iliushyk  wrote:

> These patches include fixes for issues detected by the Coverity Scan tool and 
> internal tests.
> 
> Danylo Vodopianov (26):
>   net/ntnic: fix index verification
>   net/ntnic: add thread check return code
>   net/ntnic: add return code handling
>   net/ntnic: add array index verification
>   net/ntnic: fix realloc memory leak
>   net/ntnic: fix array index verification
>   net/ntnic: add var definition transparently
>   net/ntnic: add proper var freed
>   net/ntnic: remove deadcode
>   net/ntnic: fix potentially overflow
>   net/ntnic: add null checking
>   net/ntnic: fix overflow issue
>   net/ntnic: fix untrusted loop bound
>   net/ntnic: add null checking
>   net/ntnic: move null checking
>   net/ntnic: fix var size
>   net/ntnic: fix var overflow
>   net/ntnic: remove dead code
>   net/ntnic: remove convert error func
>   net/ntnic: fix array verification
>   net/ntnic: fix memory leak
>   net/ntnic: remove unused code
>   net/ntnic: refactor RSS implementation
>   net/ntnic: fix age timeout recalculation into fpga unit
>   net/ntnic: rework age event generation
>   net/ntnic: fix group print
> 
> Oleksandr Kolomeiets (2):
>   net/ntnic: remove extra address-of operator
>   net/ntnic: remove extra check for null
> 
> Serhii Iliushyk (3):
>   net/ntnic: extend module mapping
>   net/ntnic: refactoring of the FPGA initialization
>   net/ntnic: remove tag EXPERIMENTAL
> 
>  MAINTAINERS   |   2 +-
>  .../net/ntnic/adapter/nt4ga_stat/nt4ga_stat.c |   9 -
>  drivers/net/ntnic/dbsconfig/ntnic_dbsconfig.c |  10 +-
>  drivers/net/ntnic/include/create_elements.h   |   1 -
>  drivers/net/ntnic/include/flow_api.h  |   9 -
>  drivers/net/ntnic/include/flow_api_engine.h   |   2 +
>  drivers/net/ntnic/include/hw_mod_backend.h|  16 +-
>  .../link_mgmt/link_100g/nt4ga_link_100g.c |   2 +-
>  drivers/net/ntnic/meson.build |   1 +
>  drivers/net/ntnic/nthw/core/nthw_fpga.c   |  14 +-
>  drivers/net/ntnic/nthw/flow_api/flow_api.c|  67 +-
>  drivers/net/ntnic/nthw/flow_api/flow_group.c  |  26 +
>  .../net/ntnic/nthw/flow_api/flow_hsh_cfg.c| 661 ++
>  .../net/ntnic/nthw/flow_api/flow_hsh_cfg.h|  17 +
>  .../ntnic/nthw/flow_api/hw_mod/hw_mod_flm.c   |  14 +-
>  .../ntnic/nthw/flow_api/hw_mod/hw_mod_hsh.c   |  19 +-
>  .../ntnic/nthw/flow_api/hw_mod/hw_mod_pdb.c   |  18 +-
>  .../profile_inline/flow_api_hw_db_inline.c|  29 +-
>  .../profile_inline/flow_api_profile_inline.c  | 817 +-
>  .../profile_inline/flow_api_profile_inline.h  |   4 -
>  .../ntnic/nthw/flow_filter/flow_nthw_rpp_lr.c |  10 +-
>  .../nthw/supported/nthw_fpga_mod_str_map.c|  24 +
>  drivers/net/ntnic/ntnic_ethdev.c  |  16 +-
>  drivers/net/ntnic/ntnic_filter/ntnic_filter.c | 155 ++--
>  drivers/net/ntnic/ntnic_mod_reg.h |   6 -
>  drivers/net/ntnic/ntnic_xstats/ntnic_xstats.c |  10 +-
>  26 files changed, 915 insertions(+), 1044 deletions(-)
>  create mode 100644 drivers/net/ntnic/nthw/flow_api/flow_hsh_cfg.c
>  create mode 100644 drivers/net/ntnic/nthw/flow_api/flow_hsh_cfg.h
> 


I assume that since driver was experimental, there is no point in backporting
any of this to stable.

Re: [PATCH v1 02/31] net/ntnic: add thread check return code

[Stephen Hemminger]

On Tue, 21 Jan 2025 18:07:40 +0100
Serhii Iliushyk  wrote:

> From: Danylo Vodopianov 
> 
> CI found couple coverity problems which were fixed in this commit.
> 
> CID: 448965 Error handling issues (CHECKED_RETURN).
> 
> Thread return code check was added.
> 
> Coverity issue: 448965
> Fixes: a1ba8c473f5c ("net/ntnic: add statistics poll")
> 
> Signed-off-by: Danylo Vodopianov 
> ---
>  drivers/net/ntnic/ntnic_ethdev.c | 7 +--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/ntnic/ntnic_ethdev.c 
> b/drivers/net/ntnic/ntnic_ethdev.c
> index 2a2643a106..620d023a71 100644
> --- a/drivers/net/ntnic/ntnic_ethdev.c
> +++ b/drivers/net/ntnic/ntnic_ethdev.c
> @@ -2516,8 +2516,11 @@ static int init_shutdown(void)
>   NT_LOG(DBG, NTNIC, "Starting shutdown handler");
>   kill_pmd = 0;
>   previous_handler = signal(SIGINT, signal_handler_func_int);
> - THREAD_CREATE(&shutdown_tid, shutdown_thread, NULL);
> -
> + int ret = THREAD_CREATE(&shutdown_tid, shutdown_thread, NULL);
> + if (ret != 0) {
> + NT_LOG(ERR, NTNIC, "Failed to create shutdown thread, error 
> code: %d", ret);
> + return -1;
> + }
>   /*
>* 1 time calculation of 1 sec stat update rtc cycles to prevent stat 
> poll
>* flooding by OVS from multiple virtual port threads - no need to be 
> precise

This whole init thread and shutdown logic is broken.
There are vendor NFV applications that expect to be able to do.
  
 rte_eal_init()
 signal(SIGINT, my_shutdown_func)

This kind of problem was reported years ago with the TAP device, and it
had to change to use a different signal.

[DPDK/ethdev Bug 1622] ntnic: using signals and threads

[bugzilla]

https://bugs.dpdk.org/show_bug.cgi?id=1622

Bug ID: 1622
   Summary: ntnic: using signals and threads
   Product: DPDK
   Version: 25.03
  Hardware: All
OS: All
Status: UNCONFIRMED
  Severity: major
  Priority: Normal
 Component: ethdev
  Assignee: dev@dpdk.org
  Reporter: step...@networkplumber.org
  Target Milestone: ---

No PMD should be making its own threads and must not use standard signal types.
The ntnic PMD is spawning a thread and trying to trap SIGINT.
This is broken for many reasons:
  - application may want to manage SIGINT after rte_eal_init() is called
  - there are other shutdown signals beside SIGINT (SIGTERM, SIGQUIT, ...)
  - spawning a thread outside of EAL creates resource management issues

-- 
You are receiving this mail because:
You are the assignee for the bug.

Re: [PATCH v6 2/2] dts: add flow create/delete to testpmd shell

[Dean Marx]

Got it, thanks Luca!

[PATCH] doc: update ODM dmadev guide

[Anoob Joseph]

Update ODM dmadev guide with details on ODM PF driver application. The
kernel PF driver module is being removed and so the steps specific to
that is also removed.

Signed-off-by: Anoob Joseph 
---
 doc/guides/dmadevs/odm.rst | 39 ++
 1 file changed, 14 insertions(+), 25 deletions(-)

diff --git a/doc/guides/dmadevs/odm.rst b/doc/guides/dmadevs/odm.rst
index cc2ae95275..f4be87a966 100644
--- a/doc/guides/dmadevs/odm.rst
+++ b/doc/guides/dmadevs/odm.rst
@@ -13,18 +13,14 @@ ODM DMA device can support up to 32 queues and 16 VFs.
 Device Setup
 
 
-ODM DMA device is initialized by kernel PF driver.
-The PF kernel driver is part of Marvell software packages for Odyssey.
+The ODM DMA device is initialized by the ODM PF driver application, which is
+responsible for configuring the device and creating VFs.
 
-Kernel module can be inserted as in below example::
+The driver can be downloaded from GitHub using the following link
+``_
 
-   sudo insmod odyssey_odm.ko
-
-ODM DMA device can support up to 16 VFs::
-
-   sudo echo 16 > /sys/bus/pci/devices/\:08\:00.0/sriov_numvfs
-
-Above command creates 16 VFs with 2 queues each.
+Instructions for enabling the ODM PF driver are provided in the repository's
+README file.
 
 The ``dpdk-devbind.py`` script, included with DPDK,
 can be used to show the presence of supported hardware.
@@ -59,18 +55,19 @@ for details on operation enqueue and submission API usage.
 Performance Tuning Parameters
 ~
 
-To achieve higher performance, DMA device needs to be tuned
-using PF kernel driver module parameters.
+To achieve higher performance, the DMA device needs to be tuned using the
+parameters provided by the PF driver application.
 
-Following options are exposed by kernel PF driver via devlink interface
-for tuning performance.
+The following configuration parameters are exposed by the ODM PF driver
+application
 
-``eng_sel``
+``ENG_SEL``
 
   ODM DMA device has 2 engines internally. Engine to queue mapping is decided
-  by a hardware register which can be configured as below::
+  by a hardware register which can be configured by updating the configuration
+  file::
 
- /sbin/devlink dev param set pci/:08:00.0 name eng_sel value 
3435973836 cmode runtime
+ ENG_SEL="0x"
 
   Each bit in the register corresponds to one queue.
   Each queue would be associated with one engine.
@@ -81,11 +78,3 @@ for tuning performance.
   ``1100 1100 1100 1100 1100 1100 1100 1100``
   which allows for alternate engines to be used with alternate VFs
   (assuming the system has 16 VFs with 2 queues each).
-
-``max_load_request``
-
-  Specifies maximum outstanding load requests on internal bus.
-  Values can range from 1 to 512.
-  Set to 512 for maximum requests in flight.::
-
- /sbin/devlink dev param set pci/:08:00.0 name max_load_request value 
512 cmode runtime
-- 
2.45.2

[PATCH] net/mlx5: fix GRE open matcher does not match packet

[Maayan Kashani]

The previous GRE fix was meant to handle HWS behavior
and changed the behavior for SWS causing degradation.

After the previous fix, in case of an empty GRE match,
GRE mask was initialized with protocol full mask,
Instead of the empty mask with protocol mask equals zero.

The Current fix handles the SWS case as before the GRE fix,
and HWS according to the latest fix.
Also, combined common logic.

Fixes: 25ab2cbba31d ("net/mlx5: fix GRE flow item translation for root table")
Cc: sta...@dpdk.org
Signed-off-by: Maayan Kashani 
Acked-by: Bing Zhao 
---
 drivers/net/mlx5/mlx5_flow_dv.c | 27 +++
 1 file changed, 15 insertions(+), 12 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 41ebe0b61a4..633c41e358b 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -9839,23 +9839,26 @@ flow_dv_translate_item_gre(void *key, const struct 
rte_flow_item *item,
} gre_crks_rsvd0_ver_m, gre_crks_rsvd0_ver_v;
uint16_t protocol_m, protocol_v;
 
-   if (key_type & MLX5_SET_MATCHER_M) {
+   /* Common logic to SWS/HWS */
+   if (key_type & MLX5_SET_MATCHER_M)
MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol, 0xff);
+   else
+   MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol,
+   IPPROTO_GRE);
+   /* HWS mask logic only */
+   if (key_type & MLX5_SET_MATCHER_HS_M) {
if (!gre_m)
gre_m = &rte_flow_item_gre_mask;
gre_v = gre_m;
-   } else {
-   MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol,
-IPPROTO_GRE);
-   if (!gre_v) {
-   gre_v = &empty_gre;
-   gre_m = &empty_gre;
-   } else if (!gre_m) {
-   gre_m = &rte_flow_item_gre_mask;
-   }
-   if (key_type == MLX5_SET_MATCHER_HS_V)
-   gre_m = gre_v;
+   } else if (!gre_v) {
+   gre_v = &empty_gre;
+   gre_m = &empty_gre;
+   } else if (!gre_m) {
+   gre_m = &rte_flow_item_gre_mask;
}
+   /* SWS logic only */
+   if (key_type & MLX5_SET_MATCHER_SW_M)
+   gre_v = gre_m;
gre_crks_rsvd0_ver_m.value = rte_be_to_cpu_16(gre_m->c_rsvd0_ver);
gre_crks_rsvd0_ver_v.value = rte_be_to_cpu_16(gre_v->c_rsvd0_ver);
MLX5_SET(fte_match_set_misc, misc_v, gre_c_present,
-- 
2.21.0

[PATCH v2 1/2] test: disable non-C11 atomic tests for MSVC

[Andre Muezerie]

In general, non-C11 atomics are not to be used with MSVC.
This patch skips the non-C11 atomic tests when using MSVC.

Signed-off-by: Andre Muezerie 
---
 app/test/test_atomic.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/app/test/test_atomic.c b/app/test/test_atomic.c
index db07159e81..3f26ce88d9 100644
--- a/app/test/test_atomic.c
+++ b/app/test/test_atomic.c
@@ -20,6 +20,7 @@
 
 #include "test.h"
 
+#ifndef RTE_TOOLCHAIN_MSVC
 /*
  * Atomic Variables
  * 
@@ -632,3 +633,4 @@ test_atomic(void)
return 0;
 }
 REGISTER_FAST_TEST(atomic_autotest, false, true, test_atomic);
+#endif /* RTE_TOOLCHAIN_MSVC */
-- 
2.47.2.vfs.0.1

[PATCH v2 0/2] enable build of lib/stack when using MSVC

[Andre Muezerie]

MSVC does not support inline assembly, which is used by the
implementation of rte_atomic128_cmp_exchange and is needed
by the C11 flavor of lib/stack.

A special implementation of rte_atomic128_cmp_exchange
compatible with MSVC is added to rte_stack_lf_c11.h. It uses an
intrinsic function when using MSVC, and inline assembly when other
compilers are used.

Existing atomic tests (which are not C11 compatible) are
now skipped when using MSVC.

v2:
 * Added MSVC compatible implementation of rte_atomic128_cmp_exchange
 * Skipped non-C11 atomic tests when using MSVC

Andre Muezerie (2):
  test: disable non-C11 atomic tests for MSVC
  stack: enable build with MSVC

 app/test/test_atomic.c   |  2 ++
 lib/stack/meson.build|  6 
 lib/stack/rte_stack_lf_c11.h | 62 ++--
 3 files changed, 62 insertions(+), 8 deletions(-)

--
2.47.2.vfs.0.1

[PATCH v2 2/2] stack: enable build with MSVC

[Andre Muezerie]

An implementation compatible with MSVC is provided for
atomic128_cmp_exchange in rte_stack_lf_c11.h.

Now that the issues preventing the code needed to build lib/stack
have been addressed, it can be enabled so that it also gets built
when using the MSVC compiler.

Signed-off-by: Andre Muezerie 
---
 lib/stack/meson.build|  6 
 lib/stack/rte_stack_lf_c11.h | 62 ++--
 2 files changed, 60 insertions(+), 8 deletions(-)

diff --git a/lib/stack/meson.build b/lib/stack/meson.build
index 7631a14784..18177a742f 100644
--- a/lib/stack/meson.build
+++ b/lib/stack/meson.build
@@ -1,12 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2019 Intel Corporation
 
-if is_ms_compiler
-build = false
-reason = 'not supported building with Visual Studio Toolset'
-subdir_done()
-endif
-
 sources = files('rte_stack.c', 'rte_stack_std.c', 'rte_stack_lf.c')
 headers = files('rte_stack.h')
 # subheaders, not for direct inclusion by apps
diff --git a/lib/stack/rte_stack_lf_c11.h b/lib/stack/rte_stack_lf_c11.h
index 60d46e963b..79d40486f9 100644
--- a/lib/stack/rte_stack_lf_c11.h
+++ b/lib/stack/rte_stack_lf_c11.h
@@ -8,6 +8,64 @@
 #include 
 #include 
 
+/**
+ * The maximum lock-free data size that can be manipulated atomically using C11
+ * standard is limited to 8 bytes.
+ *
+ * This implementation for __rte_atomic128_cmp_exchange operates on 16-byte
+ * data types and is made available here so that it can be used without the
+ * need to unnecessarily expose other non-C11 atomics present in
+ * rte_atomic_64.h.
+ */
+#ifdef RTE_TOOLCHAIN_MSVC
+static inline int
+__rte_atomic128_cmp_exchange(rte_int128_t *dst,
+rte_int128_t *exp,
+const rte_int128_t *src,
+unsigned int weak,
+int success,
+int failure)
+{
+   return (int)_InterlockedCompareExchange128(
+   (int64_t volatile *) dst,
+   src->val[1], /* exchange high */
+   src->val[0], /* exchange low */
+   (int64_t *) exp /* comparand result */
+   );
+}
+#else
+static inline int
+__rte_atomic128_cmp_exchange(rte_int128_t *dst,
+rte_int128_t *exp,
+const rte_int128_t *src,
+unsigned int weak,
+int success,
+int failure)
+{
+   RTE_SET_USED(weak);
+   RTE_SET_USED(success);
+   RTE_SET_USED(failure);
+   uint8_t res;
+
+   asm volatile (
+ MPLOCKED
+ "cmpxchg16b %[dst];"
+ " sete %[res]"
+ : [dst] "=m" (dst->val[0]),
+   "=a" (exp->val[0]),
+   "=d" (exp->val[1]),
+   [res] "=r" (res)
+ : "b" (src->val[0]),
+   "c" (src->val[1]),
+   "a" (exp->val[0]),
+   "d" (exp->val[1]),
+   "m" (dst->val[0])
+ : "memory");
+
+   return res;
+}
+#endif /* RTE_TOOLCHAIN_MSVC */
+
 static __rte_always_inline unsigned int
 __rte_stack_lf_count(struct rte_stack *s)
 {
@@ -55,7 +113,7 @@ __rte_stack_lf_push_elems(struct rte_stack_lf_list *list,
/* Use the release memmodel to ensure the writes to the LF LIFO
 * elements are visible before the head pointer write.
 */
-   success = rte_atomic128_cmp_exchange(
+   success = __rte_atomic128_cmp_exchange(
(rte_int128_t *)&list->head,
(rte_int128_t *)&old_head,
(rte_int128_t *)&new_head,
@@ -155,7 +213,7 @@ __rte_stack_lf_pop_elems(struct rte_stack_lf_list *list,
 * length is visible before the head update, but
 * acquire semantics on the length update is enough.
 */
-   success = rte_atomic128_cmp_exchange(
+   success = __rte_atomic128_cmp_exchange(
(rte_int128_t *)&list->head,
(rte_int128_t *)&old_head,
(rte_int128_t *)&new_head,
-- 
2.47.2.vfs.0.1

Re: [PATCH 1/3] lib/eal: add rte_atomic128_cmp_exchange compatible with MSVC

[Andre Muezerie]

On Fri, Jan 24, 2025 at 03:27:06PM +0100, David Marchand wrote:
> On Tue, Dec 10, 2024 at 5:33 PM Andre Muezerie
>  wrote:
> >
> > MSVC does not support inline assembly, which is used by the
> > implementation of rte_atomic128_cmp_exchange and is needed
> > by lib/stack.
> >
> > Error printed by MSVC:
> >
> > stack_rte_stack_lf.c.obj : error LNK2019:
> > unresolved external symbol rte_atomic128_cmp_exchange referenced
> > in function __rte_stack_lf_push_elems
> >
> > Fix is to provide an implementation for rte_atomic128_cmp_exchange
> > which uses an intrinsic function, which is used when compiling with
> > MSVC. For other compilers the existing implementation continues to
> > be used.
> >
> > Signed-off-by: Andre Muezerie 
> > ---
> >  lib/eal/x86/include/rte_atomic.h|  4 ++--
> >  lib/eal/x86/include/rte_atomic_64.h | 18 ++
> >  2 files changed, 20 insertions(+), 2 deletions(-)
> >
> > diff --git a/lib/eal/x86/include/rte_atomic.h 
> > b/lib/eal/x86/include/rte_atomic.h
> > index c72c47c83e..e8e0e4c33c 100644
> > --- a/lib/eal/x86/include/rte_atomic.h
> > +++ b/lib/eal/x86/include/rte_atomic.h
> > @@ -288,12 +288,12 @@ static inline int 
> > rte_atomic32_dec_and_test(rte_atomic32_t *v)
> >
> >  #endif
> >
> > +#endif /* RTE_TOOLCHAIN_MSVC */
> > +
> >  #ifdef RTE_ARCH_I686
> >  #include "rte_atomic_32.h"
> >  #else
> >  #include "rte_atomic_64.h"
> >  #endif
> >
> > -#endif
> > -
> 
> This partially reverts 27da6a123414 ("eal: hide legacy atomics API for MSVC").
> It would be better to implement an equivalent to
> rte_atomic128_cmp_exchange in the DPDK "new" stdatomic API
> (rte_stdatomic.h).
> 
> 
> -- 
> David Marchand

Thanks for calling that out. After looking at the past commits I got a better 
understanding
of the reasons the atomic-related code is laid out the way it is, and I agree 
that we should
attempt to follow the same guidelines.

For that reason I changed the approach taken in the v2 I sent out today.

Let me know your thoughts.
--
Andre Muezerie

Re: [PATCH v6 00/25] Reduce code duplication across Intel NIC drivers

[Bruce Richardson]

On Tue, Jan 28, 2025 at 07:20:43AM +, Shetty, Praveen wrote:
> 
> This patchset attempts to reduce the amount of code duplication across a 
> number of Intel NIC drivers, specifically: ixgbe, i40e, iavf, and ice.
> As part of this process of deduplication, and in preparation for further 
> cleanup work, it moves all existing Intel drivers to a net/intel directory, 
> with a "common" subdirectory being created there for the code shared between 
> drivers.
> 
> The first deduplication patch extracts a function from the Rx side, otherwise 
> the majority of the changes are on the Tx side, leading to a converged Tx 
> queue structure across the 4 drivers, and a large number of common functions.
> 

> --
> 
> Series looks good to me - great work Bruce!
> Series-Acked-by:  Praveen Shetty 
> 
Thanks for giving it a a review.

Series applied to dpdk-next-net-intel.

/Bruce

[PATCH v3 2/2] net/af_xdp: Refactor af_xdp_tx_zc()

[Ariel Otilibili]

Both legs of the loop share the same logic: either to go out of the
function, or to fall through to the next instructions.

For that, maybe_cpy_pkt() is introduced.

Bugzilla ID: 1440
Signed-off-by: Ariel Otilibili 
---
 drivers/net/af_xdp/rte_eth_af_xdp.c | 131 ++--
 1 file changed, 84 insertions(+), 47 deletions(-)

diff --git a/drivers/net/af_xdp/rte_eth_af_xdp.c 
b/drivers/net/af_xdp/rte_eth_af_xdp.c
index 092bcb73aa0a..efe23283fb52 100644
--- a/drivers/net/af_xdp/rte_eth_af_xdp.c
+++ b/drivers/net/af_xdp/rte_eth_af_xdp.c
@@ -536,13 +536,80 @@ kick_tx(struct pkt_tx_queue *txq, struct xsk_ring_cons 
*cq)
}
 }
 
+static inline uint64_t
+update_addr(struct rte_mbuf *mbuf, struct xsk_umem_info *umem)
+{
+   return (uint64_t)mbuf - (uint64_t)umem->buffer
+   - umem->mb_pool->header_size;
+}
+
+static inline uint64_t
+update_offset(struct rte_mbuf *mbuf, struct xsk_umem_info *umem)
+{
+   uint64_t offset;
+
+   offset = rte_pktmbuf_mtod(mbuf, uint64_t) - (uint64_t)mbuf
+   + umem->mb_pool->header_size;
+   return offset << XSK_UNALIGNED_BUF_OFFSET_SHIFT;
+}
+
+static struct rte_mbuf *
+maybe_kick_tx(struct pkt_tx_queue *txq, uint32_t *idx_tx, struct rte_mbuf 
*mbuf)
+{
+   struct rte_mbuf *ret = mbuf;
+
+   if (!xsk_ring_prod__reserve(&txq->tx, 1, idx_tx)) {
+   kick_tx(txq, &txq->pair->cq);
+   if (!xsk_ring_prod__reserve(&txq->tx, 1, idx_tx))
+   ret = NULL;
+   }
+
+   return ret;
+}
+
+static struct rte_mbuf *
+maybe_create_mbuf(struct pkt_tx_queue *txq, uint32_t *idx_tx,
+ struct xsk_umem_info *umem) {
+   struct rte_mbuf *local_mbuf = rte_pktmbuf_alloc(umem->mb_pool);
+
+   if (local_mbuf == NULL)
+   goto out;
+
+   if (!xsk_ring_prod__reserve(&txq->tx, 1, idx_tx)) {
+   rte_pktmbuf_free(local_mbuf);
+   local_mbuf = NULL;
+   goto out;
+   }
+
+out:
+   return local_mbuf;
+}
+
+static void
+maybe_cpy_pkt(bool is_mbuf_equal, struct xsk_umem_info *umem,
+ uint64_t addr_plus_offset, struct rte_mbuf *mbuf,
+ struct xdp_desc *desc)
+{
+   void *pkt;
+
+   if(is_mbuf_equal)
+   goto out;
+
+   pkt = xsk_umem__get_data(umem->buffer, addr_plus_offset);
+   rte_memcpy(pkt, rte_pktmbuf_mtod(mbuf, void *), desc->len);
+   rte_pktmbuf_free(mbuf);
+
+out:
+   return;
+}
+
 #if defined(XDP_UMEM_UNALIGNED_CHUNK_FLAG)
 static uint16_t
 af_xdp_tx_zc(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 {
struct pkt_tx_queue *txq = queue;
struct xsk_umem_info *umem = txq->umem;
-   struct rte_mbuf *mbuf;
+   struct rte_mbuf *mbuf, *local_mbuf;
unsigned long tx_bytes = 0;
int i;
uint32_t idx_tx;
@@ -551,61 +618,31 @@ af_xdp_tx_zc(void *queue, struct rte_mbuf **bufs, 
uint16_t nb_pkts)
uint64_t addr, offset;
struct xsk_ring_cons *cq = &txq->pair->cq;
uint32_t free_thresh = cq->size >> 1;
+   bool is_true;
 
if (xsk_cons_nb_avail(cq, free_thresh) >= free_thresh)
pull_umem_cq(umem, XSK_RING_CONS__DEFAULT_NUM_DESCS, cq);
 
for (i = 0; i < nb_pkts; i++) {
mbuf = bufs[i];
+   is_true = mbuf->pool == umem->mb_pool;
 
-   if (mbuf->pool == umem->mb_pool) {
-   if (!xsk_ring_prod__reserve(&txq->tx, 1, &idx_tx)) {
-   kick_tx(txq, cq);
-   if (!xsk_ring_prod__reserve(&txq->tx, 1,
-   &idx_tx))
-   goto out;
-   }
-   desc = xsk_ring_prod__tx_desc(&txq->tx, idx_tx);
-   desc->len = mbuf->pkt_len;
-   addr = (uint64_t)mbuf - (uint64_t)umem->buffer -
-   umem->mb_pool->header_size;
-   offset = rte_pktmbuf_mtod(mbuf, uint64_t) -
-   (uint64_t)mbuf +
-   umem->mb_pool->header_size;
-   offset = offset << XSK_UNALIGNED_BUF_OFFSET_SHIFT;
-   desc->addr = addr | offset;
-   tx_bytes += desc->len;
-   count++;
-   } else {
-   struct rte_mbuf *local_mbuf =
-   rte_pktmbuf_alloc(umem->mb_pool);
-   void *pkt;
-
-   if (local_mbuf == NULL)
-   goto out;
+   if (is_true)
+   local_mbuf = maybe_kick_tx(txq, &idx_tx, mbuf);
+   else
+   local_mbuf = maybe_create_mbuf(txq, &idx_tx, umem);
 
-   if (!xsk_ring_prod__reserve(&txq->tx, 1, &idx_tx)) {
-  

[PATCH v3 0/2] Fix use after free, and refactor af_xdp_tx_zc()

[Ariel Otilibili]

Hello,

The series addresses Bugzilla ID 1440 in two steps;

1. Fix use after free.
2. Refactor af_xdp_tx_zc().

Thank you,
---
v3
* reworded commit message of patch-1
* addressed feedback of Maryam Tahhan
v2 
(https://inbox.dpdk.org/dev/20250116225151.188214-1-ariel.otilib...@6wind.com/)
* reworded commit messages
* addressed feedback from Stephen Hemminger
v1 (http://inbox.dpdk.org/dev/20250116195640.68885-1-ariel.otilib...@6wind.com/)

Ariel Otilibili (2):
  net/af_xdp: fix use after free in af_xdp_tx_zc()
  net/af_xdp: Refactor af_xdp_tx_zc()

 drivers/net/af_xdp/rte_eth_af_xdp.c | 131 ++--
 1 file changed, 84 insertions(+), 47 deletions(-)

-- 
2.30.2

[PATCH v7 2/2] dts: add flow create/delete to testpmd shell

[Dean Marx]

Add flow create/delete methods to TestPmdShell class
for initializing flow rules.

Signed-off-by: Dean Marx 
---
 dts/framework/remote_session/testpmd_shell.py | 45 +++
 1 file changed, 45 insertions(+)

diff --git a/dts/framework/remote_session/testpmd_shell.py 
b/dts/framework/remote_session/testpmd_shell.py
index 3e86f52a7e..d58e4891b0 100644
--- a/dts/framework/remote_session/testpmd_shell.py
+++ b/dts/framework/remote_session/testpmd_shell.py
@@ -1910,6 +1910,51 @@ def csum_set_hw(

{port_id}:\n{csum_output}"""
 )
 
+def flow_create(self, flow_rule: FlowRule, port_id: int) -> int:
+"""Creates a flow rule in the testpmd session.
+
+Args:
+flow_rule: :class:`FlowRule` object used for creating testpmd flow 
rule.
+port_id: Integer representing the port to use.
+verify: If :data:`True`, the output of the command is scanned
+to ensure the flow rule was created successfully.
+
+Raises:
+InteractiveCommandExecutionError: If flow rule is invalid.
+
+Returns:
+Id of created flow rule. Acts as implicit verification.
+"""
+flow_output = self.send_command(f"flow create {port_id} {flow_rule}")
+match = re.search(r"#(\d+)", flow_output)
+if match is not None:
+match_str = match.group(1)
+flow_id = int(match_str)
+return flow_id
+else:
+self._logger.debug(f"Failed to create flow rule:\n{flow_output}")
+raise InteractiveCommandExecutionError(f"Failed to create flow 
rule:\n{flow_output}")
+
+def flow_delete(self, flow_id: int, port_id: int, verify: bool = True) -> 
None:
+"""Deletes the specified flow rule from the testpmd session.
+
+Args:
+flow_id: ID of the flow to remove.
+port_id: Integer representing the port to use.
+verify: If :data:`True`, the output of the command is scanned
+to ensure the flow rule was deleted successfully.
+
+Raises:
+InteractiveCommandExecutionError: If flow rule is invalid.
+"""
+flow_output = self.send_command(f"flow destroy {port_id} rule 
{flow_id}")
+if verify:
+if "destroyed" not in flow_output:
+self._logger.debug(f"Failed to delete flow 
rule:\n{flow_output}")
+raise InteractiveCommandExecutionError(
+f"Failed to delete flow rule:\n{flow_output}"
+)
+
 @requires_started_ports
 @requires_stopped_ports
 def set_port_mtu(self, port_id: int, mtu: int, verify: bool = True) -> 
None:
-- 
2.47.0

[PATCH v7 1/2] dts: add flow rule dataclass to testpmd shell

[Dean Marx]

Add dataclass for passing in flow rule creation arguments, as well as a
__str__ method for converting to a sendable testpmd command.

Signed-off-by: Dean Marx 
---
 dts/framework/remote_session/testpmd_shell.py | 52 ++-
 1 file changed, 51 insertions(+), 1 deletion(-)

diff --git a/dts/framework/remote_session/testpmd_shell.py 
b/dts/framework/remote_session/testpmd_shell.py
index 72f81e2c99..3e86f52a7e 100644
--- a/dts/framework/remote_session/testpmd_shell.py
+++ b/dts/framework/remote_session/testpmd_shell.py
@@ -22,7 +22,15 @@
 from enum import Flag, auto
 from os import environ
 from pathlib import PurePath
-from typing import TYPE_CHECKING, Any, ClassVar, Concatenate, ParamSpec, 
TypeAlias
+from typing import (
+TYPE_CHECKING,
+Any,
+ClassVar,
+Concatenate,
+Literal,
+ParamSpec,
+TypeAlias,
+)
 
 if TYPE_CHECKING or environ.get("DTS_DOC_BUILD"):
 from enum import Enum as NoAliasEnum
@@ -705,6 +713,48 @@ class TestPmdPortStats(TextParser):
 tx_bps: int = field(metadata=TextParser.find_int(r"Tx-bps:\s+(\d+)"))
 
 
+@dataclass(kw_only=True)
+class FlowRule:
+"""Class representation of flow rule parameters.
+
+This class represents the parameters of any flow rule as per the
+following pattern:
+
+[group {group_id}] [priority {level}] [ingress] [egress]
+[user_id {user_id}] pattern {item} [/ {item} [...]] / end
+actions {action} [/ {action} [...]] / end
+"""
+
+#:
+group_id: int | None = None
+#:
+priority_level: int | None = None
+#:
+direction: Literal["ingress", "egress"]
+#:
+user_id: int | None = None
+#:
+pattern: list[str]
+#:
+actions: list[str]
+
+def __str__(self) -> str:
+"""Returns the string representation of this instance."""
+ret = ""
+pattern = " / ".join(self.pattern)
+action = " / ".join(self.actions)
+if self.group_id is not None:
+ret += f"group {self.group_id} "
+if self.priority_level is not None:
+ret += f"priority {self.priority_level} "
+ret += f"{self.direction} "
+if self.user_id is not None:
+ret += f"user_id {self.user_id} "
+ret += f"pattern {pattern} / end "
+ret += f"actions {action} / end"
+return ret
+
+
 class PacketOffloadFlag(Flag):
 """Flag representing the Packet Offload Features Flags in DPDK.
 
-- 
2.47.0

Re: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device parameters for hot plug events

[Stephen Hemminger]

On Wed, 29 Jan 2025 00:10:12 +
Long Li  wrote:

> > Subject: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device parameters for
> > hot plug events
> > 
> > On Mon, 27 Jan 2025 17:35:06 -0800
> > lon...@linuxonhyperv.com wrote:
> >   
> > > @@ -1409,9 +1424,6 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
> > >   ret_stop = hn_dev_stop(eth_dev);
> > >   hn_dev_close(eth_dev);
> > >
> > > - free(hv->vf_devargs);
> > > - hv->vf_devargs = NULL;
> > > -
> > >   hn_detach(hv);
> > >   hn_chim_uninit(eth_dev);
> > >   rte_vmbus_chan_close(hv->channels[0]);
> > > @@ -1423,6 +1435,61 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
> > >   return ret_stop;
> > >  }
> > >
> > > +static int populate_cache_list(void)
> > > +{
> > > + int ret;
> > > + struct rte_devargs *da;
> > > +
> > > + rte_spinlock_lock(&netvsc_lock);
> > > + da_cache_usage++;
> > > + if (da_cache_usage > 1) {
> > > + ret = 0;
> > > + goto out;
> > > + }
> > > +
> > > + LIST_INIT(&da_cache_list);
> > > + RTE_EAL_DEVARGS_FOREACH("pci", da) {
> > > + struct da_cache *cache;
> > > +
> > > + cache = rte_zmalloc("NETVSC-HOTADD", sizeof(*cache),  
> > rte_mem_page_size());  
> > > + if (!cache) {
> > > + ret = -ENOMEM;
> > > + goto out;
> > > + }
> > > +
> > > + strncpy(cache->name, da->name, sizeof(da->name));
> > > + cache->drv_str = strdup(da->drv_str);
> > > + if (!cache->drv_str) {
> > > + rte_free(cache);
> > > + ret = -ENOMEM;
> > > + goto out;
> > > + }
> > > +
> > > + LIST_INSERT_HEAD(&da_cache_list, cache, list);
> > > + }  
> > 
> > Why do you need to cache entry to be page aligned, that seems unnecessary
> > wasteful?  
> 
> You are right it doesn't need to. I'm removing the alignment.
> 
> > Why does it need to be huge pages? versus normal malloc?  
> 
> I thought it would make it easier to debug memory leaks through EAL. But 
> normal malloc is fine because this data is not shared between 
> primary/secondary.
> 
> > The string is coming from malloc (strdup) so it can't be used by secondary 
> > process.
> > 
> > Since you are allocating a devargs cache entry, you could just as well use 
> > flexible
> > array where entry was like:
> > struct da_cache {
> > LIST_ENTRY(da_cache) list;
> > char name[RTE_DEV_NAME_MAX_LEN];
> > char drv_str[];
> > };  
> 
> This is difficult as I don't know how big it is for drv_str[].

That is why the malloc adds extra space for drv_str, see below.
Flexible array allow one array at end of structure 
> > 
> > 
> > cache = malloc(sizeof(*cache) + strlen(da->drv_str) + 1); ...
> > strcpy(cache->drv_str, da->drv_str);  
> 
> Another approach is to modify EAL to never delete driver arguments when a 
> device is removed. i.e., It doesn't call rte_devargs_remove() on device 
> removal, instead keep those devargs for the lifetime of the process. Do you 
> think this is a better approach? This will save work if other drivers want to 
> cache devargs list for device hot plug events.

Not sure, right now I don't think it is possible to pre-add devargs for future 
hot plug

RE: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device parameters for hot plug events

[Long Li]

> -Original Message-
> From: Stephen Hemminger 
> Sent: Tuesday, January 28, 2025 4:32 PM
> To: Long Li 
> Cc: lon...@linuxonhyperv.com; Ferruh Yigit ; Andrew
> Rybchenko ; Wei Hu ;
> dev@dpdk.org
> Subject: Re: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device parameters
> for hot plug events
> 
> On Wed, 29 Jan 2025 00:10:12 +
> Long Li  wrote:
> 
> > > Subject: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device
> > > parameters for hot plug events
> > >
> > > On Mon, 27 Jan 2025 17:35:06 -0800
> > > lon...@linuxonhyperv.com wrote:
> > >
> > > > @@ -1409,9 +1424,6 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
> > > > ret_stop = hn_dev_stop(eth_dev);
> > > > hn_dev_close(eth_dev);
> > > >
> > > > -   free(hv->vf_devargs);
> > > > -   hv->vf_devargs = NULL;
> > > > -
> > > > hn_detach(hv);
> > > > hn_chim_uninit(eth_dev);
> > > > rte_vmbus_chan_close(hv->channels[0]);
> > > > @@ -1423,6 +1435,61 @@ eth_hn_dev_uninit(struct rte_eth_dev
> *eth_dev)
> > > > return ret_stop;
> > > >  }
> > > >
> > > > +static int populate_cache_list(void) {
> > > > +   int ret;
> > > > +   struct rte_devargs *da;
> > > > +
> > > > +   rte_spinlock_lock(&netvsc_lock);
> > > > +   da_cache_usage++;
> > > > +   if (da_cache_usage > 1) {
> > > > +   ret = 0;
> > > > +   goto out;
> > > > +   }
> > > > +
> > > > +   LIST_INIT(&da_cache_list);
> > > > +   RTE_EAL_DEVARGS_FOREACH("pci", da) {
> > > > +   struct da_cache *cache;
> > > > +
> > > > +   cache = rte_zmalloc("NETVSC-HOTADD", sizeof(*cache),
> > > rte_mem_page_size());
> > > > +   if (!cache) {
> > > > +   ret = -ENOMEM;
> > > > +   goto out;
> > > > +   }
> > > > +
> > > > +   strncpy(cache->name, da->name, sizeof(da->name));
> > > > +   cache->drv_str = strdup(da->drv_str);
> > > > +   if (!cache->drv_str) {
> > > > +   rte_free(cache);
> > > > +   ret = -ENOMEM;
> > > > +   goto out;
> > > > +   }
> > > > +
> > > > +   LIST_INSERT_HEAD(&da_cache_list, cache, list);
> > > > +   }
> > >
> > > Why do you need to cache entry to be page aligned, that seems
> > > unnecessary wasteful?
> >
> > You are right it doesn't need to. I'm removing the alignment.
> >
> > > Why does it need to be huge pages? versus normal malloc?
> >
> > I thought it would make it easier to debug memory leaks through EAL. But
> normal malloc is fine because this data is not shared between 
> primary/secondary.
> >
> > > The string is coming from malloc (strdup) so it can't be used by secondary
> process.
> > >
> > > Since you are allocating a devargs cache entry, you could just as
> > > well use flexible array where entry was like:
> > > struct da_cache {
> > >   LIST_ENTRY(da_cache) list;
> > >   char name[RTE_DEV_NAME_MAX_LEN];
> > >   char drv_str[];
> > > };
> >
> > This is difficult as I don't know how big it is for drv_str[].
> 
> That is why the malloc adds extra space for drv_str, see below.
> Flexible array allow one array at end of structure
> > >
> > >
> > >   cache = malloc(sizeof(*cache) + strlen(da->drv_str) + 1); ...
> > >   strcpy(cache->drv_str, da->drv_str);
> >

Thank you, got it.

[PATCH v3 1/2] net/af_xdp: fix use after free in af_xdp_tx_zc()

[Ariel Otilibili]

tx_bytes is computed after both legs are tested. This might
produce a use after memory free.

The computation is now moved into each leg.

Bugzilla ID: 1440
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Signed-off-by: Ariel Otilibili 
---
 drivers/net/af_xdp/rte_eth_af_xdp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/af_xdp/rte_eth_af_xdp.c 
b/drivers/net/af_xdp/rte_eth_af_xdp.c
index 814398ba4b44..092bcb73aa0a 100644
--- a/drivers/net/af_xdp/rte_eth_af_xdp.c
+++ b/drivers/net/af_xdp/rte_eth_af_xdp.c
@@ -574,6 +574,7 @@ af_xdp_tx_zc(void *queue, struct rte_mbuf **bufs, uint16_t 
nb_pkts)
umem->mb_pool->header_size;
offset = offset << XSK_UNALIGNED_BUF_OFFSET_SHIFT;
desc->addr = addr | offset;
+   tx_bytes += desc->len;
count++;
} else {
struct rte_mbuf *local_mbuf =
@@ -601,11 +602,10 @@ af_xdp_tx_zc(void *queue, struct rte_mbuf **bufs, 
uint16_t nb_pkts)
desc->addr = addr | offset;
rte_memcpy(pkt, rte_pktmbuf_mtod(mbuf, void *),
desc->len);
+   tx_bytes += desc->len;
rte_pktmbuf_free(mbuf);
count++;
}
-
-   tx_bytes += mbuf->pkt_len;
}
 
 out:
-- 
2.30.2

RE: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device parameters for hot plug events

[Long Li]

> Subject: [EXTERNAL] Re: [PATCH 4/4] net/netvsc: cache device parameters for
> hot plug events
> 
> On Mon, 27 Jan 2025 17:35:06 -0800
> lon...@linuxonhyperv.com wrote:
> 
> > @@ -1409,9 +1424,6 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
> > ret_stop = hn_dev_stop(eth_dev);
> > hn_dev_close(eth_dev);
> >
> > -   free(hv->vf_devargs);
> > -   hv->vf_devargs = NULL;
> > -
> > hn_detach(hv);
> > hn_chim_uninit(eth_dev);
> > rte_vmbus_chan_close(hv->channels[0]);
> > @@ -1423,6 +1435,61 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
> > return ret_stop;
> >  }
> >
> > +static int populate_cache_list(void)
> > +{
> > +   int ret;
> > +   struct rte_devargs *da;
> > +
> > +   rte_spinlock_lock(&netvsc_lock);
> > +   da_cache_usage++;
> > +   if (da_cache_usage > 1) {
> > +   ret = 0;
> > +   goto out;
> > +   }
> > +
> > +   LIST_INIT(&da_cache_list);
> > +   RTE_EAL_DEVARGS_FOREACH("pci", da) {
> > +   struct da_cache *cache;
> > +
> > +   cache = rte_zmalloc("NETVSC-HOTADD", sizeof(*cache),
> rte_mem_page_size());
> > +   if (!cache) {
> > +   ret = -ENOMEM;
> > +   goto out;
> > +   }
> > +
> > +   strncpy(cache->name, da->name, sizeof(da->name));
> > +   cache->drv_str = strdup(da->drv_str);
> > +   if (!cache->drv_str) {
> > +   rte_free(cache);
> > +   ret = -ENOMEM;
> > +   goto out;
> > +   }
> > +
> > +   LIST_INSERT_HEAD(&da_cache_list, cache, list);
> > +   }
> 
> Why do you need to cache entry to be page aligned, that seems unnecessary
> wasteful?

You are right it doesn't need to. I'm removing the alignment.

> Why does it need to be huge pages? versus normal malloc?

I thought it would make it easier to debug memory leaks through EAL. But normal 
malloc is fine because this data is not shared between primary/secondary.

> The string is coming from malloc (strdup) so it can't be used by secondary 
> process.
> 
> Since you are allocating a devargs cache entry, you could just as well use 
> flexible
> array where entry was like:
> struct da_cache {
>   LIST_ENTRY(da_cache) list;
>   char name[RTE_DEV_NAME_MAX_LEN];
>   char drv_str[];
> };

This is difficult as I don't know how big it is for drv_str[].

> 
> 
>   cache = malloc(sizeof(*cache) + strlen(da->drv_str) + 1); ...
>   strcpy(cache->drv_str, da->drv_str);

Another approach is to modify EAL to never delete driver arguments when a 
device is removed. i.e., It doesn't call rte_devargs_remove() on device 
removal, instead keep those devargs for the lifetime of the process. Do you 
think this is a better approach? This will save work if other drivers want to 
cache devargs list for device hot plug events.

Thanks,
Long

[PATCH v7 04/15] net/xsc: add xsc dev ops to support VFIO driver

[Renyong Wan]

XSC PMD is designed to support both VFIO and private kernel drivers.
This commit add xsc dev ops to support VFIO driver.

Signed-off-by: Na Na 
Signed-off-by: Renyong Wan 

---

v7:
* Delete the packed attributes of sub structures and unions.

v6:
* Replace __rte_packed with __rte_packed_begin and __rte_packed_end.

v5:
* Using RTE_ETHER_ADDR_LEN instead of numeral 6.
* Initialize some local variables while be defined.
* Fix compilation warnings.
---
 drivers/net/xsc/meson.build |   1 +
 drivers/net/xsc/xsc_defs.h  |   8 +
 drivers/net/xsc/xsc_dev.h   |  32 ++
 drivers/net/xsc/xsc_rxtx.h  | 102 +
 drivers/net/xsc/xsc_vfio.c  | 746 
 5 files changed, 889 insertions(+)
 create mode 100644 drivers/net/xsc/xsc_rxtx.h
 create mode 100644 drivers/net/xsc/xsc_vfio.c

diff --git a/drivers/net/xsc/meson.build b/drivers/net/xsc/meson.build
index df4c8ea499..4e20b30438 100644
--- a/drivers/net/xsc/meson.build
+++ b/drivers/net/xsc/meson.build
@@ -10,4 +10,5 @@ sources = files(
 'xsc_ethdev.c',
 'xsc_dev.c',
 'xsc_vfio_mbox.c',
+'xsc_vfio.c',
 )
diff --git a/drivers/net/xsc/xsc_defs.h b/drivers/net/xsc/xsc_defs.h
index a4b36685a6..8fd59133bc 100644
--- a/drivers/net/xsc/xsc_defs.h
+++ b/drivers/net/xsc/xsc_defs.h
@@ -16,6 +16,14 @@
 
 #define XSC_VFREP_BASE_LOGICAL_PORT1081
 
+#define XSC_PF_TX_DB_ADDR  0x4802000
+#define XSC_PF_RX_DB_ADDR  0x4804000
+#define XSC_PF_CQ_DB_ADDR  0x212
+
+#define XSC_VF_RX_DB_ADDR  0x8d4
+#define XSC_VF_TX_DB_ADDR  0x8d0
+#define XSC_VF_CQ_DB_ADDR  0x8c4
+
 enum xsc_nic_mode {
XSC_NIC_MODE_LEGACY,
XSC_NIC_MODE_SWITCHDEV,
diff --git a/drivers/net/xsc/xsc_dev.h b/drivers/net/xsc/xsc_dev.h
index fb97bb2230..1274f9287e 100644
--- a/drivers/net/xsc/xsc_dev.h
+++ b/drivers/net/xsc/xsc_dev.h
@@ -14,6 +14,7 @@
 
 #include "xsc_defs.h"
 #include "xsc_log.h"
+#include "xsc_rxtx.h"
 
 #define XSC_PPH_MODE_ARG   "pph_mode"
 #define XSC_NIC_MODE_ARG   "nic_mode"
@@ -25,6 +26,18 @@
 #define XSC_DEV_PCT_IDX_INVALID0x
 #define XSC_DEV_REPR_ID_INVALID0x7FFF
 
+enum xsc_queue_type {
+   XSC_QUEUE_TYPE_RDMA_RC  = 0,
+   XSC_QUEUE_TYPE_RDMA_MAD = 1,
+   XSC_QUEUE_TYPE_RAW  = 2,
+   XSC_QUEUE_TYPE_VIRTIO_NET   = 3,
+   XSC_QUEUE_TYPE_VIRTIO_BLK   = 4,
+   XSC_QUEUE_TYPE_RAW_TPE  = 5,
+   XSC_QUEUE_TYPE_RAW_TSO  = 6,
+   XSC_QUEUE_TYPE_RAW_TX   = 7,
+   XSC_QUEUE_TYPE_INVALID  = 0xFF,
+};
+
 struct xsc_hwinfo {
uint32_t pcie_no; /* pcie number , 0 or 1 */
uint32_t func_id; /* pf glb func id */
@@ -120,6 +133,25 @@ struct xsc_dev_ops {
enum rte_pci_kernel_driver kdrv;
int (*dev_init)(struct xsc_dev *xdev);
int (*dev_close)(struct xsc_dev *xdev);
+   int (*get_mac)(struct xsc_dev *xdev, uint8_t *mac);
+   int (*set_link_up)(struct xsc_dev *xdev);
+   int (*set_link_down)(struct xsc_dev *xdev);
+   int (*link_update)(struct xsc_dev *xdev, uint8_t funcid_type, int 
wait_to_complete);
+   int (*set_mtu)(struct xsc_dev *xdev, uint16_t mtu);
+   int (*destroy_qp)(void *qp);
+   int (*destroy_cq)(void *cq);
+   int (*modify_qp_status)(struct xsc_dev *xdev,
+   uint32_t qpn, int num, int opcode);
+   int (*modify_qp_qostree)(struct xsc_dev *xdev, uint16_t qpn);
+
+   int (*rx_cq_create)(struct xsc_dev *xdev, struct xsc_rx_cq_params 
*cq_params,
+   struct xsc_rx_cq_info *cq_info);
+   int (*tx_cq_create)(struct xsc_dev *xdev, struct xsc_tx_cq_params 
*cq_params,
+   struct xsc_tx_cq_info *cq_info);
+   int (*tx_qp_create)(struct xsc_dev *xdev, struct xsc_tx_qp_params 
*qp_params,
+   struct xsc_tx_qp_info *qp_info);
+   int (*mailbox_exec)(struct xsc_dev *xdev, void *data_in,
+   int in_len, void *data_out, int out_len);
 };
 
 void xsc_dev_ops_register(struct xsc_dev_ops *new_ops);
diff --git a/drivers/net/xsc/xsc_rxtx.h b/drivers/net/xsc/xsc_rxtx.h
new file mode 100644
index 00..1a7d621a95
--- /dev/null
+++ b/drivers/net/xsc/xsc_rxtx.h
@@ -0,0 +1,102 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright 2025 Yunsilicon Technology Co., Ltd.
+ */
+
+#ifndef _XSC_RXTX_H_
+#define _XSC_RXTX_H_
+
+#include 
+
+struct __rte_packed_begin xsc_wqe_data_seg {
+   union {
+   struct {
+   uint8_t in_line:1;
+   uint8_t rsv0:7;
+   };
+   struct {
+   rte_le32_t  rsv1:1;
+   rte_le32_t  seg_len:31;
+   rte_le32_t  lkey;
+   rte_le64_t  va;
+   };
+   struct {
+   uint8

[PATCH v7 03/15] net/xsc: add xsc mailbox

[Renyong Wan]

XSC mailbox is a mechanism used for interaction between PMD and firmware.

Signed-off-by: Rong Qian 
Signed-off-by: Renyong Wan 

---

v7:
* Remove the cast of xdev->dev_priv.
* Remove the cast of malloc return value.
---
 drivers/net/xsc/meson.build |   1 +
 drivers/net/xsc/xsc_cmd.h   | 387 ++
 drivers/net/xsc/xsc_defs.h  |   2 +
 drivers/net/xsc/xsc_vfio_mbox.c | 691 
 drivers/net/xsc/xsc_vfio_mbox.h | 142 +++
 5 files changed, 1223 insertions(+)
 create mode 100644 drivers/net/xsc/xsc_cmd.h
 create mode 100644 drivers/net/xsc/xsc_vfio_mbox.c
 create mode 100644 drivers/net/xsc/xsc_vfio_mbox.h

diff --git a/drivers/net/xsc/meson.build b/drivers/net/xsc/meson.build
index 683a1f6632..df4c8ea499 100644
--- a/drivers/net/xsc/meson.build
+++ b/drivers/net/xsc/meson.build
@@ -9,4 +9,5 @@ endif
 sources = files(
 'xsc_ethdev.c',
 'xsc_dev.c',
+'xsc_vfio_mbox.c',
 )
diff --git a/drivers/net/xsc/xsc_cmd.h b/drivers/net/xsc/xsc_cmd.h
new file mode 100644
index 00..433dcd0afa
--- /dev/null
+++ b/drivers/net/xsc/xsc_cmd.h
@@ -0,0 +1,387 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright 2025 Yunsilicon Technology Co., Ltd.
+ */
+
+#ifndef _XSC_CMD_H_
+#define _XSC_CMD_H_
+
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#define XSC_BOARD_SN_LEN   32
+#define XSC_CMD_QUERY_HCA_CAP_V1   1
+
+enum xsc_cmd_opcode {
+   XSC_CMD_OP_QUERY_HCA_CAP= 0x100,
+   XSC_CMD_OP_CREATE_CQ= 0x400,
+   XSC_CMD_OP_DESTROY_CQ   = 0x401,
+   XSC_CMD_OP_CREATE_QP= 0x500,
+   XSC_CMD_OP_DESTROY_QP   = 0x501,
+   XSC_CMD_OP_RTR2RTS_QP   = 0x504,
+   XSC_CMD_OP_QP_2RST  = 0x50A,
+   XSC_CMD_OP_CREATE_MULTI_QP  = 0x515,
+   XSC_CMD_OP_MODIFY_NIC_HCA   = 0x812,
+   XSC_CMD_OP_MODIFY_RAW_QP= 0x81f,
+   XSC_CMD_OP_EXEC_NP  = 0x900,
+   XSC_CMD_OP_SET_MTU  = 0x1100,
+   XSC_CMD_OP_QUERY_ETH_MAC= 0X1101,
+   XSC_CMD_OP_MAX
+};
+
+enum xsc_cmd_status {
+   XSC_CMD_SUCC = 0,
+   XSC_CMD_FAIL,
+   XSC_CMD_TIMEOUT,
+};
+
+struct xsc_cmd_inbox_hdr {
+   rte_be16_t opcode;
+   uint8_t rsvd[4];
+   rte_be16_t ver;
+};
+
+struct xsc_cmd_outbox_hdr {
+   uint8_t status;
+   uint8_t rsvd[5];
+   rte_be16_t ver;
+};
+
+struct xsc_cmd_fw_version {
+   uint8_t major;
+   uint8_t minor;
+   rte_be16_t patch;
+   rte_be32_t tweak;
+   uint8_t extra_flag;
+   uint8_t rsv[7];
+};
+
+struct xsc_cmd_hca_cap {
+   uint8_t rsvd1[12];
+   uint8_t send_seg_num;
+   uint8_t send_wqe_shift;
+   uint8_t recv_seg_num;
+   uint8_t recv_wqe_shift;
+   uint8_t log_max_srq_sz;
+   uint8_t log_max_qp_sz;
+   uint8_t log_max_mtt;
+   uint8_t log_max_qp;
+   uint8_t log_max_strq_sz;
+   uint8_t log_max_srqs;
+   uint8_t rsvd2[2];
+   uint8_t log_max_tso;
+   uint8_t log_max_cq_sz;
+   uint8_t rsvd3;
+   uint8_t log_max_cq;
+   uint8_t log_max_eq_sz;
+   uint8_t log_max_mkey;
+   uint8_t log_max_msix;
+   uint8_t log_max_eq;
+   uint8_t max_indirection;
+   uint8_t log_max_mrw_sz;
+   uint8_t log_max_bsf_list_sz;
+   uint8_t log_max_klm_list_sz;
+   uint8_t rsvd4;
+   uint8_t log_max_ra_req_dc;
+   uint8_t rsvd5;
+   uint8_t log_max_ra_res_dc;
+   uint8_t rsvd6;
+   uint8_t log_max_ra_req_qp;
+   uint8_t log_max_qp_depth;
+   uint8_t log_max_ra_res_qp;
+   rte_be16_t max_vfs;
+   rte_be16_t raweth_qp_id_end;
+   rte_be16_t raw_tpe_qp_num;
+   rte_be16_t max_qp_count;
+   rte_be16_t raweth_qp_id_base;
+   uint8_t rsvd7;
+   uint8_t local_ca_ack_delay;
+   uint8_t max_num_eqs;
+   uint8_t num_ports;
+   uint8_t log_max_msg;
+   uint8_t mac_port;
+   rte_be16_t raweth_rss_qp_id_base;
+   rte_be16_t stat_rate_support;
+   uint8_t rsvd8[2];
+   rte_be64_t flags;
+   uint8_t rsvd9;
+   uint8_t uar_sz;
+   uint8_t rsvd10;
+   uint8_t log_pg_sz;
+   rte_be16_t bf_log_bf_reg_size;
+   rte_be16_t msix_base;
+   rte_be16_t msix_num;
+   rte_be16_t max_desc_sz_sq;
+   uint8_t rsvd11[2];
+   rte_be16_t max_desc_sz_rq;
+   uint8_t rsvd12[2];
+   rte_be16_t max_desc_sz_sq_dc;
+   uint8_t rsvd13[4];
+   rte_be16_t max_qp_mcg;
+   uint8_t rsvd14;
+   uint8_t log_max_mcg;
+   uint8_t rsvd15;
+   uint8_t log_max_pd;
+   uint8_t rsvd16;
+   uint8_t log_max_xrcd;
+   uint8_t rsvd17[40];
+   rte_be32_t uar_page_sz;
+   uint8_t rsvd18[8];
+   rte_be32_t hw_feature_flag;
+   rte_be16_t pf0_vf_funcid_base;
+   rte_be16_t pf0_vf_funcid_top;
+   rte_be16_t pf1_vf_funcid_base;
+   rte_be16_t pf1_vf_funcid_top;
+   rte_be16_t pcie0_pf_funcid_base;
+   

[PATCH v7 05/15] net/xsc: add PCT interfaces

[Renyong Wan]

PCT is the abbreviation of Packet classifier table, which is built
in NP to define behavior of various packets.

Signed-off-by: Renyong Wan 

---

v6:
* Replace __rte_packed with __rte_packed_begin and __rte_packed_end.
---
 drivers/net/xsc/meson.build |   1 +
 drivers/net/xsc/xsc_defs.h  |  29 +++
 drivers/net/xsc/xsc_dev.c   |  19 +-
 drivers/net/xsc/xsc_dev.h   |   3 +
 drivers/net/xsc/xsc_np.c| 492 
 drivers/net/xsc/xsc_np.h| 154 +++
 6 files changed, 697 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/xsc/xsc_np.c
 create mode 100644 drivers/net/xsc/xsc_np.h

diff --git a/drivers/net/xsc/meson.build b/drivers/net/xsc/meson.build
index 4e20b30438..5ee03ea835 100644
--- a/drivers/net/xsc/meson.build
+++ b/drivers/net/xsc/meson.build
@@ -11,4 +11,5 @@ sources = files(
 'xsc_dev.c',
 'xsc_vfio_mbox.c',
 'xsc_vfio.c',
+'xsc_np.c',
 )
diff --git a/drivers/net/xsc/xsc_defs.h b/drivers/net/xsc/xsc_defs.h
index 8fd59133bc..b1e37a5870 100644
--- a/drivers/net/xsc/xsc_defs.h
+++ b/drivers/net/xsc/xsc_defs.h
@@ -16,6 +16,26 @@
 
 #define XSC_VFREP_BASE_LOGICAL_PORT1081
 
+#define XSC_RSS_HASH_KEY_LEN   52
+#define XSC_RSS_HASH_BIT_IPV4_SIP  (1ULL << 0)
+#define XSC_RSS_HASH_BIT_IPV4_DIP  (1ULL << 1)
+#define XSC_RSS_HASH_BIT_IPV6_SIP  (1ULL << 2)
+#define XSC_RSS_HASH_BIT_IPV6_DIP  (1ULL << 3)
+#define XSC_RSS_HASH_BIT_IPV4_SPORT(1ULL << 4)
+#define XSC_RSS_HASH_BIT_IPV4_DPORT(1ULL << 5)
+#define XSC_RSS_HASH_BIT_IPV6_SPORT(1ULL << 6)
+#define XSC_RSS_HASH_BIT_IPV6_DPORT(1ULL << 7)
+#define XSC_RSS_HASH_BIT_TNL_ID(1ULL << 8)
+#define XSC_RSS_HASH_BIT_NXT_PRO   (1ULL << 9)
+
+#define XSC_EPAT_VLD_FLAG  (1ULL)
+#define XSC_EPAT_RX_QP_ID_OFST_FLAG(1ULL << 2)
+#define XSC_EPAT_QP_NUM_FLAG   (1ULL << 3)
+#define XSC_EPAT_RSS_EN_FLAG   (1ULL << 4)
+#define XSC_EPAT_RSS_HASH_TEMPLATE_FLAG(1ULL << 5)
+#define XSC_EPAT_RSS_HASH_FUNC_FLAG(1ULL << 6)
+#define XSC_EPAT_HAS_PPH_FLAG  (1ULL << 9)
+
 #define XSC_PF_TX_DB_ADDR  0x4802000
 #define XSC_PF_RX_DB_ADDR  0x4804000
 #define XSC_PF_CQ_DB_ADDR  0x212
@@ -38,4 +58,13 @@ enum xsc_pph_type {
XSC_UPLINK_PPH  = 0x8,
 };
 
+enum xsc_port_type {
+   XSC_PORT_TYPE_NONE = 0,
+   XSC_PORT_TYPE_UPLINK,
+   XSC_PORT_TYPE_UPLINK_BOND,
+   XSC_PORT_TYPE_PFVF,
+   XSC_PORT_TYPE_PFHPF,
+   XSC_PORT_TYPE_UNKNOWN,
+};
+
 #endif /* XSC_DEFS_H_ */
diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
index b030eb3410..8933f77b8f 100644
--- a/drivers/net/xsc/xsc_dev.c
+++ b/drivers/net/xsc/xsc_dev.c
@@ -54,8 +54,17 @@ xsc_dev_ops_register(struct xsc_dev_ops *new_ops)
 }
 
 int
-xsc_dev_close(struct xsc_dev *xdev, int __rte_unused repr_id)
+xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
+int in_len, void *data_out, int out_len)
 {
+   return xdev->dev_ops->mailbox_exec(xdev, data_in, in_len,
+  data_out, out_len);
+}
+
+int
+xsc_dev_close(struct xsc_dev *xdev, int repr_id)
+{
+   xsc_dev_clear_pct(xdev, repr_id);
return xdev->dev_ops->dev_close(xdev);
 }
 
@@ -121,6 +130,7 @@ void
 xsc_dev_uninit(struct xsc_dev *xdev)
 {
PMD_INIT_FUNC_TRACE();
+   xsc_dev_pct_uninit();
xsc_dev_close(xdev, XSC_DEV_REPR_ID_INVALID);
rte_free(xdev);
 }
@@ -159,6 +169,13 @@ xsc_dev_init(struct rte_pci_device *pci_dev, struct 
xsc_dev **xdev)
goto hwinfo_init_fail;
}
 
+   ret = xsc_dev_pct_init();
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Failed to init xsc pct");
+   ret = -EINVAL;
+   goto hwinfo_init_fail;
+   }
+
*xdev = d;
 
return 0;
diff --git a/drivers/net/xsc/xsc_dev.h b/drivers/net/xsc/xsc_dev.h
index 1274f9287e..51cf2d15fe 100644
--- a/drivers/net/xsc/xsc_dev.h
+++ b/drivers/net/xsc/xsc_dev.h
@@ -15,6 +15,7 @@
 #include "xsc_defs.h"
 #include "xsc_log.h"
 #include "xsc_rxtx.h"
+#include "xsc_np.h"
 
 #define XSC_PPH_MODE_ARG   "pph_mode"
 #define XSC_NIC_MODE_ARG   "nic_mode"
@@ -154,6 +155,8 @@ struct xsc_dev_ops {
int in_len, void *data_out, int out_len);
 };
 
+int xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
+int in_len, void *data_out, int out_len);
 void xsc_dev_ops_register(struct xsc_dev_ops *new_ops);
 int xsc_dev_init(struct rte_pci_device *pci_dev, struct xsc_dev **dev);
 void xsc_dev_uninit(struct xsc_dev *xdev);
diff --git a/drivers/net/xsc/xsc_np.c b/drivers/net/xsc/xsc_np.c
new file mode 100644
index 00..d4eb833bf6
--- /dev/null
+++ b/drivers/net/xsc/xsc_np.c
@@ -0,0 +1,492 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright 2025 Yunsilicon Technology Co., Ltd.
+ */
+
+#include 
+#include 
+
+#include "xsc_log.h"
+

[PATCH v7 06/15] net/xsc: initialize xsc representors

[Renyong Wan]

For the design of the xsc PMD, each ethdev corresponds to a representor.

Signed-off-by: Renyong Wan 
---
 drivers/net/xsc/xsc_defs.h   |  11 +++
 drivers/net/xsc/xsc_dev.c|  95 
 drivers/net/xsc/xsc_dev.h|   3 +
 drivers/net/xsc/xsc_ethdev.c | 170 +++
 drivers/net/xsc/xsc_ethdev.h |  19 
 5 files changed, 298 insertions(+)

diff --git a/drivers/net/xsc/xsc_defs.h b/drivers/net/xsc/xsc_defs.h
index b1e37a5870..111776f37e 100644
--- a/drivers/net/xsc/xsc_defs.h
+++ b/drivers/net/xsc/xsc_defs.h
@@ -6,6 +6,7 @@
 #define XSC_DEFS_H_
 
 #define XSC_PAGE_SIZE  4096
+#define XSC_PHY_PORT_NUM   1
 
 #define XSC_PCI_VENDOR_ID  0x1f67
 #define XSC_PCI_DEV_ID_MS  0x
@@ -15,6 +16,7 @@
 #define XSC_PCI_DEV_ID_MVS 0x1153
 
 #define XSC_VFREP_BASE_LOGICAL_PORT1081
+#define XSC_MAX_MAC_ADDRESSES  3
 
 #define XSC_RSS_HASH_KEY_LEN   52
 #define XSC_RSS_HASH_BIT_IPV4_SIP  (1ULL << 0)
@@ -58,6 +60,15 @@ enum xsc_pph_type {
XSC_UPLINK_PPH  = 0x8,
 };
 
+enum xsc_funcid_type {
+   XSC_FUNCID_TYPE_INVAL   = 0x0,
+   XSC_EMU_FUNCID  = 0x1,
+   XSC_PHYPORT_MAC_FUNCID  = 0x2,
+   XSC_VF_IOCTL_FUNCID = 0x3,
+   XSC_PHYPORT_LAG_FUNCID  = 0x4,
+   XSC_FUNCID_TYPE_UNKNOWN = 0x5,
+};
+
 enum xsc_port_type {
XSC_PORT_TYPE_NONE = 0,
XSC_PORT_TYPE_UPLINK,
diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
index 8933f77b8f..aaf18bf8e5 100644
--- a/drivers/net/xsc/xsc_dev.c
+++ b/drivers/net/xsc/xsc_dev.c
@@ -61,6 +61,12 @@ xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
   data_out, out_len);
 }
 
+int
+xsc_dev_get_mac(struct xsc_dev *xdev, uint8_t *mac)
+{
+   return xdev->dev_ops->get_mac(xdev, mac);
+}
+
 int
 xsc_dev_close(struct xsc_dev *xdev, int repr_id)
 {
@@ -126,6 +132,95 @@ xsc_dev_args_parse(struct xsc_dev *xdev, struct 
rte_devargs *devargs)
rte_kvargs_free(kvlist);
 }
 
+int
+xsc_dev_qp_set_id_get(struct xsc_dev *xdev, int repr_id)
+{
+   if (xsc_dev_is_vf(xdev))
+   return 0;
+
+   return (repr_id % 511 + 1);
+}
+
+static void
+xsc_repr_info_init(struct xsc_dev *xdev, struct xsc_repr_info *info,
+  enum xsc_port_type port_type,
+  enum xsc_funcid_type funcid_type, int32_t repr_id)
+{
+   int qp_set_id, logical_port;
+   struct xsc_hwinfo *hwinfo = &xdev->hwinfo;
+
+   info->repr_id = repr_id;
+   info->port_type = port_type;
+   if (port_type == XSC_PORT_TYPE_UPLINK_BOND) {
+   info->pf_bond = 1;
+   info->funcid = XSC_PHYPORT_LAG_FUNCID << 14;
+   } else if (port_type == XSC_PORT_TYPE_UPLINK) {
+   info->pf_bond = -1;
+   info->funcid = funcid_type << 14;
+   } else if (port_type == XSC_PORT_TYPE_PFVF) {
+   info->funcid = funcid_type << 14;
+   }
+
+   qp_set_id = xsc_dev_qp_set_id_get(xdev, repr_id);
+   if (xsc_dev_is_vf(xdev))
+   logical_port = xdev->hwinfo.func_id +
+  xdev->hwinfo.funcid_to_logic_port_off;
+   else
+   logical_port = xdev->vfos_logical_in_port + qp_set_id - 1;
+
+   info->logical_port = logical_port;
+   info->local_dstinfo = logical_port;
+   info->peer_logical_port = hwinfo->mac_phy_port;
+   info->peer_dstinfo = hwinfo->mac_phy_port;
+}
+
+int
+xsc_dev_repr_ports_probe(struct xsc_dev *xdev, int nb_repr_ports, int 
max_eth_ports)
+{
+   int funcid_type;
+   struct xsc_repr_port *repr_port;
+   int i;
+
+   PMD_INIT_FUNC_TRACE();
+
+   xdev->num_repr_ports = nb_repr_ports + XSC_PHY_PORT_NUM;
+   if (xdev->num_repr_ports > max_eth_ports) {
+   PMD_DRV_LOG(ERR, "Repr ports num %u, should be less than max 
%u",
+   xdev->num_repr_ports, max_eth_ports);
+   return -EINVAL;
+   }
+
+   xdev->repr_ports = rte_zmalloc(NULL,
+  sizeof(struct xsc_repr_port) * 
xdev->num_repr_ports,
+  RTE_CACHE_LINE_SIZE);
+   if (xdev->repr_ports == NULL) {
+   PMD_DRV_LOG(ERR, "Failed to allocate memory for repr ports");
+   return -ENOMEM;
+   }
+
+   funcid_type = (xdev->devargs.nic_mode == XSC_NIC_MODE_SWITCHDEV) ?
+   XSC_VF_IOCTL_FUNCID : XSC_PHYPORT_MAC_FUNCID;
+
+   /* PF representor use the last repr_ports */
+   repr_port = &xdev->repr_ports[xdev->num_repr_ports - 1];
+   xsc_repr_info_init(xdev, &repr_port->info, XSC_PORT_TYPE_UPLINK,
+  XSC_PHYPORT_MAC_FUNCID, xdev->num_repr_ports - 1);
+   repr_port->info.ifindex = xdev->ifindex;
+   repr_port->xdev = xdev;
+   LIST_INIT(&repr_port->def_pct_list);
+
+   /* VF representor start from 0 */
+   for (i = 0;

[PATCH v7 01/15] net/xsc: add xsc PMD framework

[Renyong Wan]

Add xsc PMD framework, doc and build infrastructure, supporting
PCI probe.

Signed-off-by: Renyong Wan 

---

v7:
* Change the name of Wanrenyong to Renyong Wan.

v6:
* Delete the included header file net/if.h.
* Rebase the main branch to merge release notes.
---
 .mailmap   |  5 ++
 MAINTAINERS| 10 +++
 doc/guides/nics/features/xsc.ini   |  9 +++
 doc/guides/nics/index.rst  |  1 +
 doc/guides/nics/xsc.rst| 31 +
 doc/guides/rel_notes/release_25_03.rst |  4 ++
 drivers/net/meson.build|  1 +
 drivers/net/xsc/meson.build| 11 
 drivers/net/xsc/xsc_defs.h | 15 +
 drivers/net/xsc/xsc_ethdev.c   | 88 ++
 drivers/net/xsc/xsc_ethdev.h   | 15 +
 drivers/net/xsc/xsc_log.h  | 24 +++
 12 files changed, 214 insertions(+)
 create mode 100644 doc/guides/nics/features/xsc.ini
 create mode 100644 doc/guides/nics/xsc.rst
 create mode 100644 drivers/net/xsc/meson.build
 create mode 100644 drivers/net/xsc/xsc_defs.h
 create mode 100644 drivers/net/xsc/xsc_ethdev.c
 create mode 100644 drivers/net/xsc/xsc_ethdev.h
 create mode 100644 drivers/net/xsc/xsc_log.h

diff --git a/.mailmap b/.mailmap
index ebead7f072..a0e77d45c9 100644
--- a/.mailmap
+++ b/.mailmap
@@ -371,6 +371,7 @@ Dongdong Liu 
 Dongsheng Rong 
 Dongsu Han 
 Dong Wang 
+Dongwei Xu 
 Dongyang Pan <197020...@qq.com>
 Dong Zhou  
 Don Provan 
@@ -1065,6 +1066,7 @@ Nagadheeraj Rottela 
 Naga Harish K S V 
 Naga Suresh Somarowthu 
 Nalla Pradeep 
+Na Na 
 Na Na 
 Nan Chen 
 Nandini Persad 
@@ -1273,6 +1275,7 @@ Remi Pommarel 
 Remigiusz Konca 
 Remy Horton 
 Renata Saiakhova 
+Renyong Wan 
 Reshma Pattan  
 Ricardo Roldan 
 Ricardo Salveti 
@@ -1309,6 +1312,7 @@ Ronak Doshi  
 Ron Beider 
 Ronghua Zhang 
 RongQiang Xie 
+Rong Qian 
 RongQing Li 
 Rongwei Liu 
 Rory Sexton 
@@ -1689,6 +1693,7 @@ Xiaonan Zhang 
 Xiao Wang 
 Xiaoxiao Zeng 
 Xiaoxin Peng 
+Xiaoxiong Zhang 
 Xiaoyu Min  
 Xiaoyun Li 
 Xiaoyun Wang 
diff --git a/MAINTAINERS b/MAINTAINERS
index b86cdd266b..4ee8098f10 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1075,6 +1075,16 @@ F: drivers/net/avp/
 F: doc/guides/nics/avp.rst
 F: doc/guides/nics/features/avp.ini
 
+Yunsilicon xsc
+M: Renyong Wan 
+M: Na Na 
+M: Rong Qian 
+M: Xiaoxiong Zhang 
+M: Dongwei Xu 
+F: drivers/net/xsc/
+F: doc/guides/nics/xsc.rst
+F: doc/guides/nics/features/xsc.ini
+
 ZTE zxdh - EXPERIMENTAL
 M: Junlong Wang 
 M: Lijie Shan 
diff --git a/doc/guides/nics/features/xsc.ini b/doc/guides/nics/features/xsc.ini
new file mode 100644
index 00..b5c44ce535
--- /dev/null
+++ b/doc/guides/nics/features/xsc.ini
@@ -0,0 +1,9 @@
+;
+; Supported features of the 'xsc' network poll mode driver.
+;
+; Refer to default.ini for the full list of available PMD features.
+;
+[Features]
+Linux= Y
+ARMv8= Y
+x86-64   = Y
diff --git a/doc/guides/nics/index.rst b/doc/guides/nics/index.rst
index 50688d9f64..10a2eca3b0 100644
--- a/doc/guides/nics/index.rst
+++ b/doc/guides/nics/index.rst
@@ -70,4 +70,5 @@ Network Interface Controller Drivers
 vhost
 virtio
 vmxnet3
+xsc
 zxdh
diff --git a/doc/guides/nics/xsc.rst b/doc/guides/nics/xsc.rst
new file mode 100644
index 00..8e189db541
--- /dev/null
+++ b/doc/guides/nics/xsc.rst
@@ -0,0 +1,31 @@
+..  SPDX-License-Identifier: BSD-3-Clause
+Copyright 2024 Yunsilicon Technology Co., Ltd
+
+XSC Poll Mode Driver
+==
+
+The xsc PMD (**librte_net_xsc**) provides poll mode driver support for
+10/25/50/100/200 Gbps Yunsilicon metaScale Series Network Adapters.
+
+Supported NICs
+--
+
+The following Yunsilicon device models are supported by the same xsc driver:
+
+  - metaScale-200S
+  - metaScale-200
+  - metaScale-100Q
+  - metaScale-50
+
+Prerequisites
+--
+
+- Follow the DPDK :ref:`Getting Started Guide for Linux ` to setup 
the basic DPDK environment.
+
+- Learning about Yunsilicon metaScale Series NICs using
+  ``_.
+
+Limitations or Known issues
+---
+32bit ARCHs are not supported.
+Windows and BSD are not supported yet.
diff --git a/doc/guides/rel_notes/release_25_03.rst 
b/doc/guides/rel_notes/release_25_03.rst
index 931fe98879..3710493aad 100644
--- a/doc/guides/rel_notes/release_25_03.rst
+++ b/doc/guides/rel_notes/release_25_03.rst
@@ -72,6 +72,10 @@ New Features
 
   * Added support for virtual function VF (ngbe_vf).
 
+* **Added Yunsilicon xsc net driver [EXPERIMENTAL].**
+
+  * Added the PMD for Yunsilicon metaScale serials NICs.
+
 
 Removed Items
 -
diff --git a/drivers/net/meson.build b/drivers/net/meson.build
index dafd637ba4..c1ca7b0b39 100644
--- a/drivers/net/meson.build
+++ b/drivers/net/meson.build
@@ -63,6 +63,7 @@ drivers = [
 'vhost',
 'virtio',
 'vmxnet3',
+'

[PATCH v7 02/15] net/xsc: add xsc device initialization

[Renyong Wan]

XSC device is a hardware abstract level device serving as a handle
to interact with hardware.

Signed-off-by: Renyong Wan 

---

v7:
* Define the size of xdev->name to PCI_PRI_STR_SIZE.

v5:
* Fix coding style issue with misspelling
* Rearrange the elements in struct xsc_hwinfo to reduce holes
---
 drivers/net/xsc/meson.build  |   1 +
 drivers/net/xsc/xsc_defs.h   |  16 
 drivers/net/xsc/xsc_dev.c| 181 +++
 drivers/net/xsc/xsc_dev.h| 131 +
 drivers/net/xsc/xsc_ethdev.c |  16 +++-
 drivers/net/xsc/xsc_ethdev.h |   3 +
 6 files changed, 347 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/xsc/xsc_dev.c
 create mode 100644 drivers/net/xsc/xsc_dev.h

diff --git a/drivers/net/xsc/meson.build b/drivers/net/xsc/meson.build
index 84a09a23de..683a1f6632 100644
--- a/drivers/net/xsc/meson.build
+++ b/drivers/net/xsc/meson.build
@@ -8,4 +8,5 @@ endif
 
 sources = files(
 'xsc_ethdev.c',
+'xsc_dev.c',
 )
diff --git a/drivers/net/xsc/xsc_defs.h b/drivers/net/xsc/xsc_defs.h
index 7c91d3443f..60244425cd 100644
--- a/drivers/net/xsc/xsc_defs.h
+++ b/drivers/net/xsc/xsc_defs.h
@@ -12,4 +12,20 @@
 #define XSC_PCI_DEV_ID_MVHVF   0x1152
 #define XSC_PCI_DEV_ID_MVS 0x1153
 
+#define XSC_VFREP_BASE_LOGICAL_PORT1081
+
+enum xsc_nic_mode {
+   XSC_NIC_MODE_LEGACY,
+   XSC_NIC_MODE_SWITCHDEV,
+   XSC_NIC_MODE_SOC,
+};
+
+enum xsc_pph_type {
+   XSC_PPH_NONE= 0,
+   XSC_RX_PPH  = 0x1,
+   XSC_TX_PPH  = 0x2,
+   XSC_VFREP_PPH   = 0x4,
+   XSC_UPLINK_PPH  = 0x8,
+};
+
 #endif /* XSC_DEFS_H_ */
diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
new file mode 100644
index 00..b030eb3410
--- /dev/null
+++ b/drivers/net/xsc/xsc_dev.c
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright 2025 Yunsilicon Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include "xsc_log.h"
+#include "xsc_defs.h"
+#include "xsc_dev.h"
+
+#define XSC_DEV_DEF_FLOW_MODE  7
+
+TAILQ_HEAD(xsc_dev_ops_list, xsc_dev_ops);
+static struct xsc_dev_ops_list dev_ops_list = 
TAILQ_HEAD_INITIALIZER(dev_ops_list);
+
+static const struct xsc_dev_ops *
+xsc_dev_ops_get(enum rte_pci_kernel_driver kdrv)
+{
+   const struct xsc_dev_ops *ops;
+
+   TAILQ_FOREACH(ops, &dev_ops_list, entry) {
+   if (ops->kdrv == kdrv)
+   return ops;
+   }
+
+   return NULL;
+}
+
+void
+xsc_dev_ops_register(struct xsc_dev_ops *new_ops)
+{
+   struct xsc_dev_ops *ops;
+
+   TAILQ_FOREACH(ops, &dev_ops_list, entry) {
+   if (ops->kdrv == new_ops->kdrv) {
+   PMD_DRV_LOG(ERR, "xsc dev ops exists, kdrv=%d", 
new_ops->kdrv);
+   return;
+   }
+   }
+
+   TAILQ_INSERT_TAIL(&dev_ops_list, new_ops, entry);
+}
+
+int
+xsc_dev_close(struct xsc_dev *xdev, int __rte_unused repr_id)
+{
+   return xdev->dev_ops->dev_close(xdev);
+}
+
+static int
+xsc_dev_alloc_vfos_info(struct xsc_dev *xdev)
+{
+   struct xsc_hwinfo *hwinfo;
+   int base_lp = 0;
+
+   if (xsc_dev_is_vf(xdev))
+   return 0;
+
+   hwinfo = &xdev->hwinfo;
+   if (hwinfo->pcie_no == 1) {
+   xdev->vfrep_offset = hwinfo->func_id -
+   hwinfo->pcie1_pf_funcid_base +
+   hwinfo->pcie0_pf_funcid_top -
+   hwinfo->pcie0_pf_funcid_base  + 1;
+   } else {
+   xdev->vfrep_offset = hwinfo->func_id - 
hwinfo->pcie0_pf_funcid_base;
+   }
+
+   base_lp = XSC_VFREP_BASE_LOGICAL_PORT;
+   if (xdev->devargs.nic_mode == XSC_NIC_MODE_LEGACY)
+   base_lp += xdev->vfrep_offset;
+   xdev->vfos_logical_in_port = base_lp;
+   return 0;
+}
+
+static void
+xsc_dev_args_parse(struct xsc_dev *xdev, struct rte_devargs *devargs)
+{
+   struct rte_kvargs *kvlist;
+   struct xsc_devargs *xdevargs = &xdev->devargs;
+   const char *tmp;
+
+   kvlist = rte_kvargs_parse(devargs->args, NULL);
+   if (kvlist == NULL)
+   return;
+
+   tmp = rte_kvargs_get(kvlist, XSC_PPH_MODE_ARG);
+   if (tmp != NULL)
+   xdevargs->pph_mode = atoi(tmp);
+   else
+   xdevargs->pph_mode = XSC_PPH_NONE;
+
+   tmp = rte_kvargs_get(kvlist, XSC_NIC_MODE_ARG);
+   if (tmp != NULL)
+   xdevargs->nic_mode = atoi(tmp);
+   else
+   xdevargs->nic_mode = XSC_NIC_MODE_LEGACY;
+
+   tmp = rte_kvargs_get(kvlist, XSC_FLOW_MODE_ARG);
+   if (tmp != NULL)
+   xdevargs->flow_mode = atoi(tmp);
+   else
+   xdevargs->flow_mode = XSC_DEV_DEF_FLOW_MODE;
+
+   rte_kvargs_free(kvlist);
+}
+
+void
+xsc_dev_uninit(struct xsc_dev *xdev)
+{
+   PMD_INIT_FUNC_TRACE();
+   xsc_dev_c

[PATCH v7 07/15] net/xsc: add ethdev configure and RSS ops

[Renyong Wan]

Implement xsc ethdev configure and RSS hash functions.

Signed-off-by: Renyong Wan 

---

v6:
* Use memcpy instead of rte_memcpy.

v5:
* Remove some unnecessary parameter checks.
---
 doc/guides/nics/features/xsc.ini |  3 ++
 drivers/net/xsc/xsc_defs.h   | 15 ++
 drivers/net/xsc/xsc_dev.c| 26 ++
 drivers/net/xsc/xsc_dev.h|  1 +
 drivers/net/xsc/xsc_ethdev.c | 82 
 drivers/net/xsc/xsc_ethdev.h |  7 +++
 6 files changed, 134 insertions(+)

diff --git a/doc/guides/nics/features/xsc.ini b/doc/guides/nics/features/xsc.ini
index b5c44ce535..bdeb7a984b 100644
--- a/doc/guides/nics/features/xsc.ini
+++ b/doc/guides/nics/features/xsc.ini
@@ -4,6 +4,9 @@
 ; Refer to default.ini for the full list of available PMD features.
 ;
 [Features]
+RSS hash = Y
+RSS key update   = Y
+RSS reta update  = Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
diff --git a/drivers/net/xsc/xsc_defs.h b/drivers/net/xsc/xsc_defs.h
index 111776f37e..c445eadca1 100644
--- a/drivers/net/xsc/xsc_defs.h
+++ b/drivers/net/xsc/xsc_defs.h
@@ -78,4 +78,19 @@ enum xsc_port_type {
XSC_PORT_TYPE_UNKNOWN,
 };
 
+enum xsc_tbm_cap {
+   XSC_TBM_CAP_HASH_PPH = 0,
+   XSC_TBM_CAP_RSS,
+   XSC_TBM_CAP_PP_BYPASS,
+   XSC_TBM_CAP_PCT_DROP_CONFIG,
+};
+
+enum xsc_rss_hf {
+   XSC_RSS_HASH_KEY_UPDATE = 0,
+   XSC_RSS_HASH_TEMP_UPDATE,
+   XSC_RSS_HASH_FUNC_UPDATE,
+   XSC_RSS_RXQ_UPDATE,
+   XSC_RSS_RXQ_DROP,
+};
+
 #endif /* XSC_DEFS_H_ */
diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
index aaf18bf8e5..74d577c7d0 100644
--- a/drivers/net/xsc/xsc_dev.c
+++ b/drivers/net/xsc/xsc_dev.c
@@ -19,6 +19,7 @@
 #include "xsc_log.h"
 #include "xsc_defs.h"
 #include "xsc_dev.h"
+#include "xsc_cmd.h"
 
 #define XSC_DEV_DEF_FLOW_MODE  7
 
@@ -74,6 +75,31 @@ xsc_dev_close(struct xsc_dev *xdev, int repr_id)
return xdev->dev_ops->dev_close(xdev);
 }
 
+int
+xsc_dev_rss_key_modify(struct xsc_dev *xdev, uint8_t *rss_key, uint8_t 
rss_key_len)
+{
+   struct xsc_cmd_modify_nic_hca_mbox_in in = {};
+   struct xsc_cmd_modify_nic_hca_mbox_out out = {};
+   uint8_t rss_caps_mask = 0;
+   int ret, key_len = 0;
+
+   in.hdr.opcode = rte_cpu_to_be_16(XSC_CMD_OP_MODIFY_NIC_HCA);
+
+   key_len = RTE_MIN(rss_key_len, XSC_RSS_HASH_KEY_LEN);
+   rte_memcpy(in.rss.hash_key, rss_key, key_len);
+   rss_caps_mask |= RTE_BIT32(XSC_RSS_HASH_KEY_UPDATE);
+
+   in.rss.caps_mask = rss_caps_mask;
+   in.rss.rss_en = 1;
+   in.nic.caps_mask = rte_cpu_to_be_16(RTE_BIT32(XSC_TBM_CAP_RSS));
+   in.nic.caps = in.nic.caps_mask;
+
+   ret = xsc_dev_mailbox_exec(xdev, &in, sizeof(in), &out, sizeof(out));
+   if (ret != 0 || out.hdr.status != 0)
+   return -1;
+   return 0;
+}
+
 static int
 xsc_dev_alloc_vfos_info(struct xsc_dev *xdev)
 {
diff --git a/drivers/net/xsc/xsc_dev.h b/drivers/net/xsc/xsc_dev.h
index d317ec9d3a..36b92a334b 100644
--- a/drivers/net/xsc/xsc_dev.h
+++ b/drivers/net/xsc/xsc_dev.h
@@ -162,6 +162,7 @@ int xsc_dev_init(struct rte_pci_device *pci_dev, struct 
xsc_dev **dev);
 void xsc_dev_uninit(struct xsc_dev *xdev);
 int xsc_dev_close(struct xsc_dev *xdev, int repr_id);
 int xsc_dev_repr_ports_probe(struct xsc_dev *xdev, int nb_repr_ports, int 
max_eth_ports);
+int xsc_dev_rss_key_modify(struct xsc_dev *xdev, uint8_t *rss_key, uint8_t 
rss_key_len);
 bool xsc_dev_is_vf(struct xsc_dev *xdev);
 int xsc_dev_qp_set_id_get(struct xsc_dev *xdev, int repr_id);
 int xsc_dev_get_mac(struct xsc_dev *xdev, uint8_t *mac);
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 223719ab90..60343f0081 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -8,6 +8,81 @@
 #include "xsc_defs.h"
 #include "xsc_ethdev.h"
 
+static int
+xsc_ethdev_rss_hash_conf_get(struct rte_eth_dev *dev,
+struct rte_eth_rss_conf *rss_conf)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+
+   if (rss_conf->rss_key != NULL && rss_conf->rss_key_len >= 
priv->rss_conf.rss_key_len)
+   memcpy(rss_conf->rss_key, priv->rss_conf.rss_key, 
priv->rss_conf.rss_key_len);
+
+   rss_conf->rss_key_len = priv->rss_conf.rss_key_len;
+   rss_conf->rss_hf = priv->rss_conf.rss_hf;
+   return 0;
+}
+
+static int
+xsc_ethdev_rss_hash_update(struct rte_eth_dev *dev,
+  struct rte_eth_rss_conf *rss_conf)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   int ret = 0;
+
+   ret = xsc_dev_rss_key_modify(priv->xdev, rss_conf->rss_key, 
rss_conf->rss_key_len);
+   if (ret == 0) {
+   memcpy(priv->rss_conf.rss_key, rss_conf->rss_key,
+  priv->rss_conf.rss_key_len);
+   priv->rss_conf.rss_key_len = rss_conf->rss_key_len;
+   priv->rss_conf.rss_hf = rss_conf->

[PATCH v7 08/15] net/xsc: add Rx and Tx queue setup

[Renyong Wan]

Implement xsc ethdev Rx and Tx queue setup functions.

Signed-off-by: Rong Qian 
Signed-off-by: Renyong Wan 

---

v7:
* Delete the packed attributes of sub structures and unions.

v6:
* Replace __rte_packed with __rte_packed_begin and __rte_packed_end.
---
 drivers/net/xsc/xsc_defs.h   |  4 ++
 drivers/net/xsc/xsc_ethdev.c | 83 
 drivers/net/xsc/xsc_rx.h | 59 +
 drivers/net/xsc/xsc_rxtx.h   | 49 +
 drivers/net/xsc/xsc_tx.h | 55 
 5 files changed, 250 insertions(+)
 create mode 100644 drivers/net/xsc/xsc_rx.h
 create mode 100644 drivers/net/xsc/xsc_tx.h

diff --git a/drivers/net/xsc/xsc_defs.h b/drivers/net/xsc/xsc_defs.h
index c445eadca1..6497b53e1e 100644
--- a/drivers/net/xsc/xsc_defs.h
+++ b/drivers/net/xsc/xsc_defs.h
@@ -38,6 +38,10 @@
 #define XSC_EPAT_RSS_HASH_FUNC_FLAG(1ULL << 6)
 #define XSC_EPAT_HAS_PPH_FLAG  (1ULL << 9)
 
+#define XSC_MAX_DESC_NUMBER1024
+#define XSC_SEND_WQE_DS3
+#define XSC_ESEG_EXTRA_DATA_SIZE   48u
+
 #define XSC_PF_TX_DB_ADDR  0x4802000
 #define XSC_PF_RX_DB_ADDR  0x4804000
 #define XSC_PF_CQ_DB_ADDR  0x212
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 60343f0081..2b0907bb2e 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -7,6 +7,8 @@
 #include "xsc_log.h"
 #include "xsc_defs.h"
 #include "xsc_ethdev.h"
+#include "xsc_rx.h"
+#include "xsc_tx.h"
 
 static int
 xsc_ethdev_rss_hash_conf_get(struct rte_eth_dev *dev,
@@ -83,6 +85,85 @@ xsc_ethdev_configure(struct rte_eth_dev *dev)
return -rte_errno;
 }
 
+static int
+xsc_ethdev_rx_queue_setup(struct rte_eth_dev *dev, uint16_t idx, uint16_t desc,
+ uint32_t socket, const struct rte_eth_rxconf *conf,
+ struct rte_mempool *mp)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_rxq_data *rxq_data = NULL;
+   uint16_t desc_n;
+   uint16_t rx_free_thresh;
+   uint64_t offloads = conf->offloads | 
dev->data->dev_conf.rxmode.offloads;
+
+   desc = (desc > XSC_MAX_DESC_NUMBER) ? XSC_MAX_DESC_NUMBER : desc;
+   desc_n = desc;
+
+   if (!rte_is_power_of_2(desc))
+   desc_n = 1 << rte_log2_u32(desc);
+
+   rxq_data = rte_malloc_socket(NULL, sizeof(*rxq_data) + desc_n * 
sizeof(struct rte_mbuf *),
+RTE_CACHE_LINE_SIZE, socket);
+   if (rxq_data == NULL) {
+   PMD_DRV_LOG(ERR, "Port %u create rxq idx %d failure",
+   dev->data->port_id, idx);
+   rte_errno = ENOMEM;
+   return -rte_errno;
+   }
+   rxq_data->idx = idx;
+   rxq_data->priv = priv;
+   (*priv->rxqs)[idx] = rxq_data;
+
+   rx_free_thresh = (conf->rx_free_thresh) ? conf->rx_free_thresh : 
XSC_RX_FREE_THRESH;
+   rxq_data->rx_free_thresh = rx_free_thresh;
+
+   rxq_data->elts = (struct rte_mbuf *(*)[desc_n])(rxq_data + 1);
+   rxq_data->mp = mp;
+   rxq_data->socket = socket;
+
+   rxq_data->csum = !!(offloads & RTE_ETH_RX_OFFLOAD_CHECKSUM);
+   rxq_data->hw_timestamp = !!(offloads & RTE_ETH_RX_OFFLOAD_TIMESTAMP);
+   rxq_data->crc_present = 0;
+
+   rxq_data->wqe_n = rte_log2_u32(desc_n);
+   rxq_data->wqe_s = desc_n;
+   rxq_data->wqe_m = desc_n - 1;
+
+   rxq_data->port_id = dev->data->port_id;
+   dev->data->rx_queues[idx] = rxq_data;
+   return 0;
+}
+
+static int
+xsc_ethdev_tx_queue_setup(struct rte_eth_dev *dev, uint16_t idx, uint16_t desc,
+ uint32_t socket, const struct rte_eth_txconf *conf)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_txq_data *txq;
+   uint16_t desc_n;
+
+   desc = (desc > XSC_MAX_DESC_NUMBER) ? XSC_MAX_DESC_NUMBER : desc;
+   desc_n = desc;
+
+   if (!rte_is_power_of_2(desc))
+   desc_n = 1 << rte_log2_u32(desc);
+
+   txq = rte_malloc_socket(NULL, sizeof(*txq) + desc_n * sizeof(struct 
rte_mbuf *),
+   RTE_CACHE_LINE_SIZE, socket);
+   txq->offloads = conf->offloads | dev->data->dev_conf.txmode.offloads;
+   txq->priv = priv;
+   txq->socket = socket;
+
+   txq->elts_n = rte_log2_u32(desc_n);
+   txq->elts_s = desc_n;
+   txq->elts_m = desc_n - 1;
+   txq->port_id = dev->data->port_id;
+   txq->idx = idx;
+
+   (*priv->txqs)[idx] = txq;
+   return 0;
+}
+
 static int
 xsc_ethdev_mac_addr_add(struct rte_eth_dev *dev, struct rte_ether_addr *mac, 
uint32_t index)
 {
@@ -111,6 +192,8 @@ xsc_ethdev_mac_addr_add(struct rte_eth_dev *dev, struct 
rte_ether_addr *mac, uin
 
 const struct eth_dev_ops xsc_eth_dev_ops = {
.dev_configure = xsc_ethdev_configure,
+   .rx_queue_setup = xsc_ethdev_rx_queue_setup,
+   .tx_queue_se

[PATCH v7 09/15] net/xsc: add ethdev start

[Renyong Wan]

Implement xsc ethdev start function.

Signed-off-by: Rong Qian 
Signed-off-by: Renyong Wan 

---

v5:
* Remove unnecessary call of rte_wmb.
---
 drivers/net/xsc/meson.build  |   2 +
 drivers/net/xsc/xsc_dev.c|  33 
 drivers/net/xsc/xsc_dev.h|   8 +
 drivers/net/xsc/xsc_ethdev.c | 172 +
 drivers/net/xsc/xsc_ethdev.h |  19 +++
 drivers/net/xsc/xsc_rx.c | 291 +++
 drivers/net/xsc/xsc_rx.h |   3 +
 drivers/net/xsc/xsc_rxtx.h   |  27 
 drivers/net/xsc/xsc_tx.c |  93 +++
 drivers/net/xsc/xsc_tx.h |   4 +
 10 files changed, 652 insertions(+)
 create mode 100644 drivers/net/xsc/xsc_rx.c
 create mode 100644 drivers/net/xsc/xsc_tx.c

diff --git a/drivers/net/xsc/meson.build b/drivers/net/xsc/meson.build
index 5ee03ea835..79664374e3 100644
--- a/drivers/net/xsc/meson.build
+++ b/drivers/net/xsc/meson.build
@@ -12,4 +12,6 @@ sources = files(
 'xsc_vfio_mbox.c',
 'xsc_vfio.c',
 'xsc_np.c',
+'xsc_rx.c',
+'xsc_tx.c',
 )
diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
index 74d577c7d0..054a08bf84 100644
--- a/drivers/net/xsc/xsc_dev.c
+++ b/drivers/net/xsc/xsc_dev.c
@@ -68,6 +68,39 @@ xsc_dev_get_mac(struct xsc_dev *xdev, uint8_t *mac)
return xdev->dev_ops->get_mac(xdev, mac);
 }
 
+int
+xsc_dev_modify_qp_status(struct xsc_dev *xdev, uint32_t qpn, int num, int 
opcode)
+{
+   return xdev->dev_ops->modify_qp_status(xdev, qpn, num, opcode);
+}
+
+int
+xsc_dev_modify_qp_qostree(struct xsc_dev *xdev, uint16_t qpn)
+{
+   return xdev->dev_ops->modify_qp_qostree(xdev, qpn);
+}
+
+int
+xsc_dev_rx_cq_create(struct xsc_dev *xdev, struct xsc_rx_cq_params *cq_params,
+struct xsc_rx_cq_info *cq_info)
+{
+   return xdev->dev_ops->rx_cq_create(xdev, cq_params, cq_info);
+}
+
+int
+xsc_dev_tx_cq_create(struct xsc_dev *xdev, struct xsc_tx_cq_params *cq_params,
+struct xsc_tx_cq_info *cq_info)
+{
+   return xdev->dev_ops->tx_cq_create(xdev, cq_params, cq_info);
+}
+
+int
+xsc_dev_tx_qp_create(struct xsc_dev *xdev, struct xsc_tx_qp_params *qp_params,
+struct xsc_tx_qp_info *qp_info)
+{
+   return xdev->dev_ops->tx_qp_create(xdev, qp_params, qp_info);
+}
+
 int
 xsc_dev_close(struct xsc_dev *xdev, int repr_id)
 {
diff --git a/drivers/net/xsc/xsc_dev.h b/drivers/net/xsc/xsc_dev.h
index 36b92a334b..fafe53f752 100644
--- a/drivers/net/xsc/xsc_dev.h
+++ b/drivers/net/xsc/xsc_dev.h
@@ -158,6 +158,14 @@ struct xsc_dev_ops {
 int xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
 int in_len, void *data_out, int out_len);
 void xsc_dev_ops_register(struct xsc_dev_ops *new_ops);
+int xsc_dev_modify_qp_status(struct xsc_dev *xdev, uint32_t qpn, int num, int 
opcode);
+int xsc_dev_modify_qp_qostree(struct xsc_dev *xdev, uint16_t qpn);
+int xsc_dev_rx_cq_create(struct xsc_dev *xdev, struct xsc_rx_cq_params 
*cq_params,
+struct xsc_rx_cq_info *cq_info);
+int xsc_dev_tx_cq_create(struct xsc_dev *xdev, struct xsc_tx_cq_params 
*cq_params,
+struct xsc_tx_cq_info *cq_info);
+int xsc_dev_tx_qp_create(struct xsc_dev *xdev, struct xsc_tx_qp_params 
*qp_params,
+struct xsc_tx_qp_info *qp_info);
 int xsc_dev_init(struct rte_pci_device *pci_dev, struct xsc_dev **dev);
 void xsc_dev_uninit(struct xsc_dev *xdev);
 int xsc_dev_close(struct xsc_dev *xdev, int repr_id);
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 2b0907bb2e..24ee1e2cdc 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -9,6 +9,8 @@
 #include "xsc_ethdev.h"
 #include "xsc_rx.h"
 #include "xsc_tx.h"
+#include "xsc_dev.h"
+#include "xsc_cmd.h"
 
 static int
 xsc_ethdev_rss_hash_conf_get(struct rte_eth_dev *dev,
@@ -85,6 +87,175 @@ xsc_ethdev_configure(struct rte_eth_dev *dev)
return -rte_errno;
 }
 
+static int
+xsc_ethdev_enable(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_hwinfo *hwinfo;
+   int peer_dstinfo = 0;
+   int peer_logicalport = 0;
+   int logical_port = 0;
+   int local_dstinfo = 0;
+   int pcie_logic_port = 0;
+   int qp_set_id;
+   int repr_id;
+   struct xsc_rxq_data *rxq = xsc_rxq_get(priv, 0);
+   uint16_t rx_qpn = (uint16_t)rxq->qpn;
+   int i, vld;
+   struct xsc_txq_data *txq;
+   struct xsc_repr_port *repr;
+   struct xsc_repr_info *repr_info;
+
+   if (priv->funcid_type != XSC_PHYPORT_MAC_FUNCID)
+   return -ENODEV;
+
+   hwinfo = &priv->xdev->hwinfo;
+   repr_id = priv->representor_id;
+   repr = &priv->xdev->repr_ports[repr_id];
+   repr_info = &repr->info;
+
+   qp_set_id = xsc_dev_qp_set_id_get(priv->xdev, repr_id);
+   logical_port = repr_info->logical_port;
+   local_dstinfo = repr_info->local_dstinfo;
+ 

[PATCH v7 13/15] net/xsc: add basic stats ops

[Renyong Wan]

Implement xsc ethdev basic statatics ops.

Signed-off-by: Renyong Wan 

---

v6:
* Remove unnecessary paren.
* Add the feature of stats per queue in xsc.ini.
---
 doc/guides/nics/features/xsc.ini |  2 +
 drivers/net/xsc/xsc_ethdev.c | 75 
 2 files changed, 77 insertions(+)

diff --git a/doc/guides/nics/features/xsc.ini b/doc/guides/nics/features/xsc.ini
index 772c6418c4..56e38a27d8 100644
--- a/doc/guides/nics/features/xsc.ini
+++ b/doc/guides/nics/features/xsc.ini
@@ -11,6 +11,8 @@ L3 checksum offload  = Y
 L4 checksum offload  = Y
 Inner L3 checksum= Y
 Inner L4 checksum= Y
+Basic stats  = Y
+Stats per queue  = Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index d344714d79..63c64ce547 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -443,6 +443,79 @@ xsc_ethdev_tx_queue_setup(struct rte_eth_dev *dev, 
uint16_t idx, uint16_t desc,
return 0;
 }
 
+static int
+xsc_ethdev_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   uint32_t rxqs_n = priv->num_rq;
+   uint32_t txqs_n = priv->num_sq;
+   uint32_t i, idx;
+   struct xsc_rxq_data *rxq;
+   struct xsc_txq_data *txq;
+
+   for (i = 0; i < rxqs_n; ++i) {
+   rxq = xsc_rxq_get(priv, i);
+   if (unlikely(rxq == NULL))
+   continue;
+
+   idx = rxq->idx;
+   if (idx < RTE_ETHDEV_QUEUE_STAT_CNTRS) {
+   stats->q_ipackets[idx] += rxq->stats.rx_pkts;
+   stats->q_ibytes[idx] += rxq->stats.rx_bytes;
+   stats->q_errors[idx] += rxq->stats.rx_errors +
+   rxq->stats.rx_nombuf;
+   }
+   stats->ipackets += rxq->stats.rx_pkts;
+   stats->ibytes += rxq->stats.rx_bytes;
+   stats->ierrors += rxq->stats.rx_errors;
+   stats->rx_nombuf += rxq->stats.rx_nombuf;
+   }
+
+   for (i = 0; i < txqs_n; ++i) {
+   txq = xsc_txq_get(priv, i);
+   if (unlikely(txq == NULL))
+   continue;
+
+   idx = txq->idx;
+   if (idx < RTE_ETHDEV_QUEUE_STAT_CNTRS) {
+   stats->q_opackets[idx] += txq->stats.tx_pkts;
+   stats->q_obytes[idx] += txq->stats.tx_bytes;
+   stats->q_errors[idx] += txq->stats.tx_errors;
+   }
+   stats->opackets += txq->stats.tx_pkts;
+   stats->obytes += txq->stats.tx_bytes;
+   stats->oerrors += txq->stats.tx_errors;
+   }
+
+   return 0;
+}
+
+static int
+xsc_ethdev_stats_reset(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   uint32_t rxqs_n = priv->num_rq;
+   uint32_t txqs_n = priv->num_sq;
+   uint32_t i;
+   struct xsc_rxq_data *rxq;
+   struct xsc_txq_data *txq;
+
+   for (i = 0; i < rxqs_n; ++i) {
+   rxq = xsc_rxq_get(priv, i);
+   if (unlikely(rxq == NULL))
+   continue;
+   memset(&rxq->stats, 0, sizeof(struct xsc_rxq_stats));
+   }
+   for (i = 0; i < txqs_n; ++i) {
+   txq = xsc_txq_get(priv, i);
+   if (unlikely(txq == NULL))
+   continue;
+   memset(&txq->stats, 0, sizeof(struct xsc_txq_stats));
+   }
+
+   return 0;
+}
+
 static int
 xsc_ethdev_mac_addr_add(struct rte_eth_dev *dev, struct rte_ether_addr *mac, 
uint32_t index)
 {
@@ -474,6 +547,8 @@ const struct eth_dev_ops xsc_eth_dev_ops = {
.dev_start = xsc_ethdev_start,
.dev_stop = xsc_ethdev_stop,
.dev_close = xsc_ethdev_close,
+   .stats_get = xsc_ethdev_stats_get,
+   .stats_reset = xsc_ethdev_stats_reset,
.rx_queue_setup = xsc_ethdev_rx_queue_setup,
.tx_queue_setup = xsc_ethdev_tx_queue_setup,
.rx_queue_release = xsc_ethdev_rxq_release,
-- 
2.25.1

[PATCH v7 11/15] net/xsc: add ethdev Rx burst

[Renyong Wan]

Implement xsc ethdev Rx burst function.

Signed-off-by: Xiaoxiong Zhang 
Signed-off-by: Renyong Wan 
---
 drivers/net/xsc/xsc_ethdev.c |   3 +
 drivers/net/xsc/xsc_rx.c | 174 +++
 drivers/net/xsc/xsc_rx.h |   1 +
 drivers/net/xsc/xsc_rxtx.h   |  13 +++
 4 files changed, 191 insertions(+)

diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 0f5a6c3ba2..ebc62fd33c 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -309,6 +309,9 @@ xsc_ethdev_start(struct rte_eth_dev *dev)
}
 
dev->data->dev_started = 1;
+
+   dev->rx_pkt_burst = xsc_rx_burst;
+
ret = xsc_ethdev_enable(dev);
 
return 0;
diff --git a/drivers/net/xsc/xsc_rx.c b/drivers/net/xsc/xsc_rx.c
index 2081f3b619..58a9cc2f26 100644
--- a/drivers/net/xsc/xsc_rx.c
+++ b/drivers/net/xsc/xsc_rx.c
@@ -13,6 +13,180 @@
 
 #define XSC_MAX_RECV_LEN 9800
 
+static inline void
+xsc_cq_to_mbuf(struct xsc_rxq_data *rxq, struct rte_mbuf *pkt,
+  volatile struct xsc_cqe *cqe)
+{
+   uint32_t rss_hash_res = 0;
+
+   pkt->port = rxq->port_id;
+   if (rxq->rss_hash) {
+   rss_hash_res = rte_be_to_cpu_32(cqe->vni);
+   if (rss_hash_res) {
+   pkt->hash.rss = rss_hash_res;
+   pkt->ol_flags |= RTE_MBUF_F_RX_RSS_HASH;
+   }
+   }
+}
+
+static inline int
+xsc_rx_poll_len(struct xsc_rxq_data *rxq, volatile struct xsc_cqe *cqe)
+{
+   int len;
+
+   do {
+   len = 0;
+   int ret;
+
+   ret = xsc_check_cqe_own(cqe, rxq->cqe_n, rxq->cq_ci);
+   if (unlikely(ret != XSC_CQE_OWNER_SW)) {
+   if (unlikely(ret == XSC_CQE_OWNER_ERR)) {
+   ++rxq->stats.rx_errors;
+   if (ret == XSC_CQE_OWNER_HW || ret == -1)
+   return 0;
+   } else {
+   return 0;
+   }
+   }
+
+   rxq->cq_ci += 1;
+   len = rte_le_to_cpu_32(cqe->msg_len);
+   return len;
+   } while (1);
+}
+
+static __rte_always_inline void
+xsc_pkt_info_sync(struct rte_mbuf *rep, struct rte_mbuf *seg)
+{
+   if (rep != NULL && seg != NULL) {
+   rep->data_len = seg->data_len;
+   rep->pkt_len = seg->pkt_len;
+   rep->data_off = seg->data_off;
+   rep->port = seg->port;
+   }
+}
+
+uint16_t
+xsc_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n)
+{
+   struct xsc_rxq_data *rxq = dpdk_rxq;
+   const uint32_t wqe_m = rxq->wqe_m;
+   const uint32_t cqe_m = rxq->cqe_m;
+   const uint32_t sge_n = rxq->sge_n;
+   struct rte_mbuf *pkt = NULL;
+   struct rte_mbuf *seg = NULL;
+   volatile struct xsc_cqe *cqe = &(*rxq->cqes)[rxq->cq_ci & cqe_m];
+   uint32_t nb_pkts = 0;
+   uint64_t nb_bytes = 0;
+   uint32_t rq_ci = rxq->rq_ci;
+   int len = 0;
+   uint32_t cq_ci_two = 0;
+   int valid_cqe_num = 0;
+   int cqe_msg_len = 0;
+   volatile struct xsc_cqe_u64 *cqe_u64 = NULL;
+   struct rte_mbuf *rep;
+
+   while (pkts_n) {
+   uint32_t idx = rq_ci & wqe_m;
+   volatile struct xsc_wqe_data_seg *wqe =
+   &((volatile struct xsc_wqe_data_seg *)rxq->wqes)[idx << 
sge_n];
+
+   seg = (*rxq->elts)[idx];
+   rte_prefetch0(cqe);
+   rte_prefetch0(wqe);
+
+   rep = rte_mbuf_raw_alloc(seg->pool);
+   if (unlikely(rep == NULL)) {
+   ++rxq->stats.rx_nombuf;
+   break;
+   }
+
+   if (!pkt) {
+   if (valid_cqe_num) {
+   cqe = cqe + 1;
+   len = cqe_msg_len;
+   valid_cqe_num = 0;
+   } else if ((rxq->cq_ci % 2 == 0) && (pkts_n > 1)) {
+   cq_ci_two = (rxq->cq_ci & rxq->cqe_m) / 2;
+   cqe_u64 = &(*rxq->cqes_u64)[cq_ci_two];
+   cqe = (volatile struct xsc_cqe *)cqe_u64;
+   len = xsc_rx_poll_len(rxq, cqe);
+   if (len > 0) {
+   cqe_msg_len = xsc_rx_poll_len(rxq, cqe 
+ 1);
+   if (cqe_msg_len > 0)
+   valid_cqe_num = 1;
+   }
+   } else {
+   cqe = &(*rxq->cqes)[rxq->cq_ci & rxq->cqe_m];
+   len = xsc_rx_poll_len(rxq, cqe);
+   }
+
+   if (!len) {
+   rte_mbuf_raw_free(rep);
+   

[PATCH v7 10/15] net/xsc: add ethdev stop and close

[Renyong Wan]

Implement xsc ethdev close and stop functions.

Signed-off-by: Renyong Wan 
---
 drivers/net/xsc/xsc_dev.c|  12 
 drivers/net/xsc/xsc_dev.h|   2 +
 drivers/net/xsc/xsc_ethdev.c | 108 +++
 drivers/net/xsc/xsc_rx.c |  47 +++
 drivers/net/xsc/xsc_rx.h |   2 +
 drivers/net/xsc/xsc_tx.c |  33 +++
 drivers/net/xsc/xsc_tx.h |   2 +
 7 files changed, 206 insertions(+)

diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
index 054a08bf84..350a1fbc70 100644
--- a/drivers/net/xsc/xsc_dev.c
+++ b/drivers/net/xsc/xsc_dev.c
@@ -68,6 +68,18 @@ xsc_dev_get_mac(struct xsc_dev *xdev, uint8_t *mac)
return xdev->dev_ops->get_mac(xdev, mac);
 }
 
+int
+xsc_dev_destroy_qp(struct xsc_dev *xdev, void *qp)
+{
+   return xdev->dev_ops->destroy_qp(qp);
+}
+
+int
+xsc_dev_destroy_cq(struct xsc_dev *xdev, void *cq)
+{
+   return xdev->dev_ops->destroy_cq(cq);
+}
+
 int
 xsc_dev_modify_qp_status(struct xsc_dev *xdev, uint32_t qpn, int num, int 
opcode)
 {
diff --git a/drivers/net/xsc/xsc_dev.h b/drivers/net/xsc/xsc_dev.h
index fafe53f752..c0a4d57af7 100644
--- a/drivers/net/xsc/xsc_dev.h
+++ b/drivers/net/xsc/xsc_dev.h
@@ -158,6 +158,8 @@ struct xsc_dev_ops {
 int xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
 int in_len, void *data_out, int out_len);
 void xsc_dev_ops_register(struct xsc_dev_ops *new_ops);
+int xsc_dev_destroy_qp(struct xsc_dev *xdev, void *qp);
+int xsc_dev_destroy_cq(struct xsc_dev *xdev, void *cq);
 int xsc_dev_modify_qp_status(struct xsc_dev *xdev, uint32_t qpn, int num, int 
opcode);
 int xsc_dev_modify_qp_qostree(struct xsc_dev *xdev, uint16_t qpn);
 int xsc_dev_rx_cq_create(struct xsc_dev *xdev, struct xsc_rx_cq_params 
*cq_params,
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 24ee1e2cdc..0f5a6c3ba2 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -87,6 +87,44 @@ xsc_ethdev_configure(struct rte_eth_dev *dev)
return -rte_errno;
 }
 
+static void
+xsc_ethdev_txq_release(struct rte_eth_dev *dev, uint16_t idx)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_txq_data *txq_data = xsc_txq_get(priv, idx);
+
+   if (txq_data == NULL)
+   return;
+
+   xsc_dev_set_qpsetid(priv->xdev, txq_data->qpn, 0);
+   xsc_txq_obj_release(priv->xdev, txq_data);
+   rte_free(txq_data->fcqs);
+   txq_data->fcqs = NULL;
+   xsc_txq_elts_free(txq_data);
+   rte_free(txq_data);
+   (*priv->txqs)[idx] = NULL;
+
+   dev->data->tx_queues[idx] = NULL;
+   dev->data->tx_queue_state[idx] = RTE_ETH_QUEUE_STATE_STOPPED;
+}
+
+static void
+xsc_ethdev_rxq_release(struct rte_eth_dev *dev, uint16_t idx)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_rxq_data *rxq_data = xsc_rxq_get(priv, idx);
+
+   if (rxq_data == NULL)
+   return;
+   xsc_rxq_rss_obj_release(priv->xdev, rxq_data);
+   xsc_rxq_elts_free(rxq_data);
+   rte_free(rxq_data);
+   (*priv->rxqs)[idx] = NULL;
+
+   dev->data->rx_queues[idx] = NULL;
+   dev->data->rx_queue_state[idx] = RTE_ETH_QUEUE_STATE_STOPPED;
+}
+
 static int
 xsc_ethdev_enable(struct rte_eth_dev *dev)
 {
@@ -149,6 +187,30 @@ xsc_ethdev_enable(struct rte_eth_dev *dev)
return 0;
 }
 
+static void
+xsc_rxq_stop(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   uint16_t i;
+
+   for (i = 0; i != priv->num_rq; ++i)
+   xsc_ethdev_rxq_release(dev, i);
+   priv->rxqs = NULL;
+   priv->flags &= ~XSC_FLAG_RX_QUEUE_INIT;
+}
+
+static void
+xsc_txq_stop(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   uint16_t i;
+
+   for (i = 0; i != priv->num_sq; ++i)
+   xsc_ethdev_txq_release(dev, i);
+   priv->txqs = NULL;
+   priv->flags &= ~XSC_FLAG_TX_QUEUE_INIT;
+}
+
 static int
 xsc_txq_start(struct xsc_ethdev_priv *priv)
 {
@@ -253,9 +315,51 @@ xsc_ethdev_start(struct rte_eth_dev *dev)
 
 error:
dev->data->dev_started = 0;
+   xsc_txq_stop(dev);
+   xsc_rxq_stop(dev);
return -rte_errno;
 }
 
+static int
+xsc_ethdev_stop(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   uint16_t i;
+
+   PMD_DRV_LOG(DEBUG, "Port %u stopping", dev->data->port_id);
+   dev->data->dev_started = 0;
+   dev->rx_pkt_burst = rte_eth_pkt_burst_dummy;
+   dev->tx_pkt_burst = rte_eth_pkt_burst_dummy;
+   rte_wmb();
+
+   rte_delay_us_sleep(1000 * priv->num_rq);
+   for (i = 0; i < priv->num_rq; ++i)
+   dev->data->rx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
+   for (i = 0; i < priv->num_sq; ++i)
+   dev->data->tx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
+
+   return 0;
+}
+
+static int

[PATCH v7 14/15] net/xsc: add ethdev infos get

[Renyong Wan]

Implement xsc ethdev information get ops.

Signed-off-by: Renyong Wan 
---
 drivers/net/xsc/xsc_ethdev.c | 61 
 1 file changed, 61 insertions(+)

diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 63c64ce547..5d548adb2d 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -364,6 +364,66 @@ xsc_ethdev_close(struct rte_eth_dev *dev)
return 0;
 }
 
+static uint64_t
+xsc_get_rx_queue_offloads(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_dev_config *config = &priv->config;
+   uint64_t offloads = 0;
+
+   if (config->hw_csum)
+   offloads |= (RTE_ETH_RX_OFFLOAD_IPV4_CKSUM |
+RTE_ETH_RX_OFFLOAD_UDP_CKSUM |
+RTE_ETH_RX_OFFLOAD_TCP_CKSUM);
+
+   return offloads;
+}
+
+static uint64_t
+xsc_get_tx_port_offloads(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   uint64_t offloads = 0;
+   struct xsc_dev_config *config = &priv->config;
+
+   if (config->hw_csum)
+   offloads |= (RTE_ETH_TX_OFFLOAD_IPV4_CKSUM |
+RTE_ETH_TX_OFFLOAD_UDP_CKSUM |
+RTE_ETH_TX_OFFLOAD_TCP_CKSUM);
+   if (config->tso)
+   offloads |= RTE_ETH_TX_OFFLOAD_TCP_TSO;
+   return offloads;
+}
+
+static int
+xsc_ethdev_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *info)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+
+   info->min_rx_bufsize = 64;
+   info->max_rx_pktlen = 65536;
+   info->max_lro_pkt_size = 0;
+   info->max_rx_queues = 256;
+   info->max_tx_queues = 1024;
+   info->rx_desc_lim.nb_max = 4096;
+   info->rx_desc_lim.nb_min = 16;
+   info->tx_desc_lim.nb_max = 8192;
+   info->tx_desc_lim.nb_min = 128;
+
+   info->rx_queue_offload_capa = xsc_get_rx_queue_offloads(dev);
+   info->rx_offload_capa = info->rx_queue_offload_capa;
+   info->tx_offload_capa = xsc_get_tx_port_offloads(dev);
+
+   info->if_index = priv->ifindex;
+   info->speed_capa = priv->xdev->link_speed_capa;
+   info->hash_key_size = XSC_RSS_HASH_KEY_LEN;
+   info->tx_desc_lim.nb_seg_max = 8;
+   info->tx_desc_lim.nb_mtu_seg_max = 8;
+   info->switch_info.name = dev->data->name;
+   info->switch_info.port_id = priv->representor_id;
+   return 0;
+}
+
 static int
 xsc_ethdev_rx_queue_setup(struct rte_eth_dev *dev, uint16_t idx, uint16_t desc,
  uint32_t socket, const struct rte_eth_rxconf *conf,
@@ -549,6 +609,7 @@ const struct eth_dev_ops xsc_eth_dev_ops = {
.dev_close = xsc_ethdev_close,
.stats_get = xsc_ethdev_stats_get,
.stats_reset = xsc_ethdev_stats_reset,
+   .dev_infos_get = xsc_ethdev_infos_get,
.rx_queue_setup = xsc_ethdev_rx_queue_setup,
.tx_queue_setup = xsc_ethdev_tx_queue_setup,
.rx_queue_release = xsc_ethdev_rxq_release,
-- 
2.25.1

[PATCH v7 12/15] net/xsc: add ethdev Tx burst

[Renyong Wan]

Implement xsc ethdev Tx burst function.

Signed-off-by: Dongwei Xu 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/xsc.ini |   4 +
 drivers/net/xsc/xsc_ethdev.c |   1 +
 drivers/net/xsc/xsc_tx.c | 228 +++
 drivers/net/xsc/xsc_tx.h |   1 +
 4 files changed, 234 insertions(+)

diff --git a/doc/guides/nics/features/xsc.ini b/doc/guides/nics/features/xsc.ini
index bdeb7a984b..772c6418c4 100644
--- a/doc/guides/nics/features/xsc.ini
+++ b/doc/guides/nics/features/xsc.ini
@@ -7,6 +7,10 @@
 RSS hash = Y
 RSS key update   = Y
 RSS reta update  = Y
+L3 checksum offload  = Y
+L4 checksum offload  = Y
+Inner L3 checksum= Y
+Inner L4 checksum= Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index ebc62fd33c..d344714d79 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -311,6 +311,7 @@ xsc_ethdev_start(struct rte_eth_dev *dev)
dev->data->dev_started = 1;
 
dev->rx_pkt_burst = xsc_rx_burst;
+   dev->tx_pkt_burst = xsc_tx_burst;
 
ret = xsc_ethdev_enable(dev);
 
diff --git a/drivers/net/xsc/xsc_tx.c b/drivers/net/xsc/xsc_tx.c
index 56daf6b4c6..406fa95381 100644
--- a/drivers/net/xsc/xsc_tx.c
+++ b/drivers/net/xsc/xsc_tx.c
@@ -124,3 +124,231 @@ xsc_txq_elts_free(struct xsc_txq_data *txq_data)
}
PMD_DRV_LOG(DEBUG, "Port %u txq %u free elts", txq_data->port_id, 
txq_data->idx);
 }
+
+static __rte_always_inline void
+xsc_tx_elts_flush(struct xsc_txq_data *__rte_restrict txq, uint16_t tail)
+{
+   uint16_t elts_n = tail - txq->elts_tail;
+   uint32_t free_n;
+
+   do {
+   free_n = txq->elts_s - (txq->elts_tail & txq->elts_m);
+   free_n = RTE_MIN(free_n, elts_n);
+   rte_pktmbuf_free_bulk(&txq->elts[txq->elts_tail & txq->elts_m], 
free_n);
+   txq->elts_tail += free_n;
+   elts_n -= free_n;
+   } while (elts_n > 0);
+}
+
+static void
+xsc_tx_cqes_handle(struct xsc_txq_data *__rte_restrict txq)
+{
+   uint32_t count = XSC_TX_COMP_CQE_HANDLE_MAX;
+   volatile struct xsc_cqe *last_cqe = NULL;
+   volatile struct xsc_cqe *cqe;
+   bool doorbell = false;
+   int ret;
+   uint16_t tail;
+
+   do {
+   cqe = &txq->cqes[txq->cq_ci & txq->cqe_m];
+   ret = xsc_check_cqe_own(cqe, txq->cqe_n, txq->cq_ci);
+   if (unlikely(ret != XSC_CQE_OWNER_SW)) {
+   if (likely(ret != XSC_CQE_OWNER_ERR))
+   /* No new CQEs in completion queue. */
+   break;
+   doorbell = true;
+   ++txq->cq_ci;
+   txq->cq_pi = txq->cq_ci;
+   last_cqe = NULL;
+   ++txq->stats.tx_errors;
+   continue;
+   }
+
+   doorbell = true;
+   ++txq->cq_ci;
+   last_cqe = cqe;
+   } while (--count > 0);
+
+   if (likely(doorbell)) {
+   union xsc_cq_doorbell cq_db = {
+   .cq_data = 0
+   };
+   cq_db.next_cid = txq->cq_ci;
+   cq_db.cq_num = txq->cqn;
+
+   /* Ring doorbell */
+   rte_write32(rte_cpu_to_le_32(cq_db.cq_data), txq->cq_db);
+
+   /* Release completed elts */
+   if (likely(last_cqe != NULL)) {
+   txq->wqe_pi = rte_le_to_cpu_16(last_cqe->wqe_id) >> 
txq->wqe_ds_n;
+   tail = txq->fcqs[(txq->cq_ci - 1) & txq->cqe_m];
+   if (likely(tail != txq->elts_tail))
+   xsc_tx_elts_flush(txq, tail);
+   }
+   }
+}
+
+static __rte_always_inline void
+xsc_tx_wqe_ctrl_seg_init(struct xsc_txq_data *__rte_restrict txq,
+struct rte_mbuf *__rte_restrict mbuf,
+struct xsc_wqe *__rte_restrict wqe)
+{
+   struct xsc_send_wqe_ctrl_seg *cs = &wqe->cseg;
+   int i = 0;
+   int ds_max = (1 << txq->wqe_ds_n) - 1;
+
+   cs->msg_opcode = XSC_OPCODE_RAW;
+   cs->wqe_id = rte_cpu_to_le_16(txq->wqe_ci << txq->wqe_ds_n);
+   cs->has_pph = 0;
+   /* Clear dseg's seg len */
+   if (cs->ds_data_num > 1 && cs->ds_data_num <= ds_max) {
+   for (i = 1; i < cs->ds_data_num; i++)
+   wqe->dseg[i].seg_len = 0;
+   }
+
+   cs->ds_data_num = mbuf->nb_segs;
+   if (mbuf->ol_flags & RTE_MBUF_F_TX_IP_CKSUM)
+   cs->csum_en = 0x2;
+   else
+   cs->csum_en = 0;
+
+   if (txq->tso_en == 1 && (mbuf->ol_flags & RTE_MBUF_F_TX_TCP_SEG)) {
+   cs->has_pph = 0;
+   cs->so_type = 1;
+   cs->so_hdr_len = mbuf->l2_len + mbuf->l3_len + mbuf->l4_len;
+   cs-

[PATCH v7 15/15] net/xsc: add ethdev link and MTU ops

[Renyong Wan]

Implement xsc ethdev link and MTU ops.

Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/xsc.ini |  1 +
 drivers/net/xsc/xsc_dev.c| 33 ++
 drivers/net/xsc/xsc_dev.h|  4 +++
 drivers/net/xsc/xsc_ethdev.c | 60 
 4 files changed, 98 insertions(+)

diff --git a/doc/guides/nics/features/xsc.ini b/doc/guides/nics/features/xsc.ini
index 56e38a27d8..8d5c292b7f 100644
--- a/doc/guides/nics/features/xsc.ini
+++ b/doc/guides/nics/features/xsc.ini
@@ -4,6 +4,7 @@
 ; Refer to default.ini for the full list of available PMD features.
 ;
 [Features]
+MTU update   = Y
 RSS hash = Y
 RSS key update   = Y
 RSS reta update  = Y
diff --git a/drivers/net/xsc/xsc_dev.c b/drivers/net/xsc/xsc_dev.c
index 350a1fbc70..c836f2f35a 100644
--- a/drivers/net/xsc/xsc_dev.c
+++ b/drivers/net/xsc/xsc_dev.c
@@ -62,6 +62,39 @@ xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
   data_out, out_len);
 }
 
+int
+xsc_dev_set_link_up(struct xsc_dev *xdev)
+{
+   if (xdev->dev_ops->set_link_up == NULL)
+   return -ENOTSUP;
+
+   return xdev->dev_ops->set_link_up(xdev);
+}
+
+int
+xsc_dev_set_link_down(struct xsc_dev *xdev)
+{
+   if (xdev->dev_ops->set_link_down == NULL)
+   return -ENOTSUP;
+
+   return xdev->dev_ops->set_link_down(xdev);
+}
+
+int
+xsc_dev_link_update(struct xsc_dev *xdev, uint8_t funcid_type, int 
wait_to_complete)
+{
+   if (xdev->dev_ops->link_update == NULL)
+   return -ENOTSUP;
+
+   return xdev->dev_ops->link_update(xdev, funcid_type, wait_to_complete);
+}
+
+int
+xsc_dev_set_mtu(struct xsc_dev *xdev, uint16_t mtu)
+{
+   return xdev->dev_ops->set_mtu(xdev, mtu);
+}
+
 int
 xsc_dev_get_mac(struct xsc_dev *xdev, uint8_t *mac)
 {
diff --git a/drivers/net/xsc/xsc_dev.h b/drivers/net/xsc/xsc_dev.h
index c0a4d57af7..cc5fa8aad8 100644
--- a/drivers/net/xsc/xsc_dev.h
+++ b/drivers/net/xsc/xsc_dev.h
@@ -158,6 +158,9 @@ struct xsc_dev_ops {
 int xsc_dev_mailbox_exec(struct xsc_dev *xdev, void *data_in,
 int in_len, void *data_out, int out_len);
 void xsc_dev_ops_register(struct xsc_dev_ops *new_ops);
+int xsc_dev_set_link_up(struct xsc_dev *xdev);
+int xsc_dev_set_link_down(struct xsc_dev *xde);
+int xsc_dev_link_update(struct xsc_dev *xdev, uint8_t funcid_type, int 
wait_to_complete);
 int xsc_dev_destroy_qp(struct xsc_dev *xdev, void *qp);
 int xsc_dev_destroy_cq(struct xsc_dev *xdev, void *cq);
 int xsc_dev_modify_qp_status(struct xsc_dev *xdev, uint32_t qpn, int num, int 
opcode);
@@ -175,6 +178,7 @@ int xsc_dev_repr_ports_probe(struct xsc_dev *xdev, int 
nb_repr_ports, int max_et
 int xsc_dev_rss_key_modify(struct xsc_dev *xdev, uint8_t *rss_key, uint8_t 
rss_key_len);
 bool xsc_dev_is_vf(struct xsc_dev *xdev);
 int xsc_dev_qp_set_id_get(struct xsc_dev *xdev, int repr_id);
+int xsc_dev_set_mtu(struct xsc_dev *xdev, uint16_t mtu);
 int xsc_dev_get_mac(struct xsc_dev *xdev, uint8_t *mac);
 
 #endif /* _XSC_DEV_H_ */
diff --git a/drivers/net/xsc/xsc_ethdev.c b/drivers/net/xsc/xsc_ethdev.c
index 5d548adb2d..b9675a5218 100644
--- a/drivers/net/xsc/xsc_ethdev.c
+++ b/drivers/net/xsc/xsc_ethdev.c
@@ -364,6 +364,41 @@ xsc_ethdev_close(struct rte_eth_dev *dev)
return 0;
 }
 
+static int
+xsc_ethdev_set_link_up(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_dev *xdev = priv->xdev;
+
+   return xsc_dev_set_link_up(xdev);
+}
+
+static int
+xsc_ethdev_set_link_down(struct rte_eth_dev *dev)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_dev *xdev = priv->xdev;
+
+   return xsc_dev_set_link_down(xdev);
+}
+
+static int
+xsc_ethdev_link_update(struct rte_eth_dev *dev,
+  int wait_to_complete)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   struct xsc_dev *xdev = priv->xdev;
+   int ret = 0;
+
+   ret = xsc_dev_link_update(xdev, priv->funcid_type, wait_to_complete);
+   if (ret == 0) {
+   dev->data->dev_link = xdev->pf_dev_link;
+   dev->data->dev_link.link_autoneg = 
!(dev->data->dev_conf.link_speeds &
+ RTE_ETH_LINK_SPEED_FIXED);
+   }
+   return ret;
+}
+
 static uint64_t
 xsc_get_rx_queue_offloads(struct rte_eth_dev *dev)
 {
@@ -503,6 +538,27 @@ xsc_ethdev_tx_queue_setup(struct rte_eth_dev *dev, 
uint16_t idx, uint16_t desc,
return 0;
 }
 
+static int
+xsc_ethdev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
+{
+   struct xsc_ethdev_priv *priv = TO_XSC_ETHDEV_PRIV(dev);
+   int ret = 0;
+
+   if (priv->eth_type != RTE_ETH_REPRESENTOR_PF) {
+   priv->mtu = mtu;
+   return 0;
+   }
+
+   ret = xsc_dev_set_mtu(priv->xdev, mtu);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Mtu set to %u failure", mtu);
+  

[PATCH v7 00/15] XSC PMD for Yunsilicon NICs

[Renyong Wan]

This xsc PMD (**librte_net_xsc**) provides poll mode driver for 
Yunsilicon metaScale serials NICs.

Features:
-
- MTU update
- TSO
- RSS hash
- RSS key update
- RSS reta update
- L3 checksum offload
- L4 checksum offload
- Inner L3 checksum
- Inner L4 checksum
- Basic stats
- Stats per queue

Support NICs:
-
- metaScale-200S   Single QSFP56 Port 200GE SmartNIC
- metaScale-200Quad QSFP28 Ports 100GE SmartNIC
- metaScale-50 Dual QSFP28 Port 25GE SmartNIC
- metaScale-100Q   Quad QSFP28 Port 25GE SmartNIC

---

v7:
* Remove the packed attributes of sub structures and unions.
* Remove the cast of xdev->dev_priv.
* Remove the cast of malloc return value.
* Define the size of xdev->name to PCI_PRI_STR_SIZE.
* Change the name of Wanrenyong to Renyong Wan

v6:
* Remove unnecessary paren.
* Add the feature of stats per queue in xsc.ini.
* Use memcpy instead of rte_memcpy in non critical path.
* Rebase from the main branch and merge release notes.
* Replace __rte_packed with __rte_packed_begin and __rte_packed_end.

v5:
* fix compilation errors.
* fix coding style issue with misspelling.
* remove some unnecessary parameter checks.
* remove unnecessary call of rte_wmb.
* Rearrange elements in structure to avoid holes.

v4:
* Based on the review comments from previous versions, reconstruct the xsc PMD 
to eliminate
  the dependency on rdma core library and proprietary kernel driver, while 
adding support for
  the vfio kernel driver.

v3:
* fix compilation errors

v2:
* fix checkpatch warnings and errors

---
Renyong Wan (15):
  net/xsc: add xsc PMD framework
  net/xsc: add xsc device initialization
  net/xsc: add xsc mailbox
  net/xsc: add xsc dev ops to support VFIO driver
  net/xsc: add PCT interfaces
  net/xsc: initialize xsc representors
  net/xsc: add ethdev configure and RSS ops
  net/xsc: add Rx and Tx queue setup
  net/xsc: add ethdev start
  net/xsc: add ethdev stop and close
  net/xsc: add ethdev Rx burst
  net/xsc: add ethdev Tx burst
  net/xsc: add basic stats ops
  net/xsc: add ethdev infos get
  net/xsc: add ethdev link and MTU ops

 .mailmap   |   5 +
 MAINTAINERS|  10 +
 doc/guides/nics/features/xsc.ini   |  19 +
 doc/guides/nics/index.rst  |   1 +
 doc/guides/nics/xsc.rst|  31 +
 doc/guides/rel_notes/release_25_03.rst |   4 +
 drivers/net/meson.build|   1 +
 drivers/net/xsc/meson.build|  17 +
 drivers/net/xsc/xsc_cmd.h  | 387 +++
 drivers/net/xsc/xsc_defs.h | 100 +++
 drivers/net/xsc/xsc_dev.c  | 397 +++
 drivers/net/xsc/xsc_dev.h  | 184 +
 drivers/net/xsc/xsc_ethdev.c   | 917 +
 drivers/net/xsc/xsc_ethdev.h   |  63 ++
 drivers/net/xsc/xsc_log.h  |  24 +
 drivers/net/xsc/xsc_np.c   | 492 +
 drivers/net/xsc/xsc_np.h   | 154 +
 drivers/net/xsc/xsc_rx.c   | 512 ++
 drivers/net/xsc/xsc_rx.h   |  65 ++
 drivers/net/xsc/xsc_rxtx.h | 191 +
 drivers/net/xsc/xsc_tx.c   | 354 ++
 drivers/net/xsc/xsc_tx.h   |  62 ++
 drivers/net/xsc/xsc_vfio.c | 746 
 drivers/net/xsc/xsc_vfio_mbox.c| 691 +++
 drivers/net/xsc/xsc_vfio_mbox.h| 142 
 25 files changed, 5569 insertions(+)
 create mode 100644 doc/guides/nics/features/xsc.ini
 create mode 100644 doc/guides/nics/xsc.rst
 create mode 100644 drivers/net/xsc/meson.build
 create mode 100644 drivers/net/xsc/xsc_cmd.h
 create mode 100644 drivers/net/xsc/xsc_defs.h
 create mode 100644 drivers/net/xsc/xsc_dev.c
 create mode 100644 drivers/net/xsc/xsc_dev.h
 create mode 100644 drivers/net/xsc/xsc_ethdev.c
 create mode 100644 drivers/net/xsc/xsc_ethdev.h
 create mode 100644 drivers/net/xsc/xsc_log.h
 create mode 100644 drivers/net/xsc/xsc_np.c
 create mode 100644 drivers/net/xsc/xsc_np.h
 create mode 100644 drivers/net/xsc/xsc_rx.c
 create mode 100644 drivers/net/xsc/xsc_rx.h
 create mode 100644 drivers/net/xsc/xsc_rxtx.h
 create mode 100644 drivers/net/xsc/xsc_tx.c
 create mode 100644 drivers/net/xsc/xsc_tx.h
 create mode 100644 drivers/net/xsc/xsc_vfio.c
 create mode 100644 drivers/net/xsc/xsc_vfio_mbox.c
 create mode 100644 drivers/net/xsc/xsc_vfio_mbox.h

-- 
2.25.1

Re: [PATCH] devtool: fix falsely reporting from checkpatch

[David Marchand]

Hi,

On Mon, Jan 20, 2025 at 12:27 PM WanRenyong  wrote:
>
> When executes the check_packed_attributes function in checkpatch,
> if __rte_packed_begin or __rte_packed_end appear in the context
> of a patch file, there may be a situation where the counts of
> __rte_packed_begin and __rte_packed_end do not match, causing
> checkpatch to return a failure.
> This patch fixes this issue by only counting the lines in the
> patch file that start with a + and include either
> __rte_packed_begin or __rte_packed_end.
>
> Signed-off-by: WanRenyong 

Thanks for proposing this fix.

When sending such fixes, don't forget to add a Fixes: and Cc:
author/maintainers by using --cc-cmd devtools/get-maintainers.sh


Adding Akhil, André and Thomas in the loop.

I also had some concern about the check:
https://inbox.dpdk.org/dev/CAJFAV8w=s1L-WYk+Qv-B+Mn6eAwKrB=GTz6hU--ZoLrJsz7=d...@mail.gmail.com/

But I merged the check untouched as nobody else seemed to object.
Checkpatch warnings are known to have false positive and such false
positives are "filtered" by subtree maintainers.


> ---
>  devtools/checkpatches.sh | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/devtools/checkpatches.sh b/devtools/checkpatches.sh
> index 003bb49e04..2e228b7f92 100755
> --- a/devtools/checkpatches.sh
> +++ b/devtools/checkpatches.sh
> @@ -384,8 +384,8 @@ check_packed_attributes() { # 
> res=1
> fi
>
> -   begin_count=$(grep '__rte_packed_begin' "$1" | wc -l)
> -   end_count=$(grep '__rte_packed_end' "$1" | wc -l)
> +   begin_count=$(grep -E '^\+.*__rte_packed_begin' "$1" | wc -l)
> +   end_count=$(grep -E '^\+.*__rte_packed_end' "$1" | wc -l)
> if [ $begin_count != $end_count ]; then
> echo "__rte_packed_begin and __rte_packed_end should always 
> be used in pairs."
> res=1

I don't think this change is any better.

There is a good chance that a patch touching just a first line of a
structure definition won't come along a line touching the last line of
the struct.
My suggestion (if we want to avoid those non useful warning) would be
to just remove the counting stuff in the check.

Opinions?


-- 
David Marchand

DPDK Release Status Meeting 2025-01-28

[Mcnamara, John]

Release status meeting minutes 2025-01-28
=

Agenda:
* Release Dates
* Subtrees
* Roadmaps
* LTS
* Defects
* Opens

Participants:
* ARM
* Debian
* Intel
* Microsoft
* Marvell
* Nvidia
* Red Hat


Release Dates
-

The following are the proposed working dates for 25.03:

- Proposal deadline (RFC/v1 patches): 4 January 2025
- API freeze (-rc1): 7 February 2025
- PMD features freeze (-rc2): 28 February 2025
- Builtin applications features freeze (-rc3): 7 March 2025
- Release: 26 March 2025


https://core.dpdk.org/roadmap/


Subtrees


* next-net
  * No update.

* next-net-intel
  * Several patches merged on intel tree
  * Patch to deduplicate Intel driver code
  * ~ 15 patches on the backlog
  * Expecting a large patch to the igc/e1000

* next-net-mlx
  * Some patches merged.
  * No big changes in this release.

* next-net-mvl
  * Some patches merged.
  * 5 patches remaining.

* next-eventdev
  * Some patches merged.
  * 5 patches remaining.

* next-baseband
  * 1 series submitting. Awaiting review.

* next-virtio
  * 1 PR merged to fix regression.
  * Backlog of patches to review.

* next-crypto
  * ~30 patches on backlog
  * Series on ZTE driver
  * Issue with checkpatch

* next-dts
  * 34 patches in backlog. Around the same number merged.
  * Issue with Python 3.13

* main
  * The meeting has moved to a US friendly time slot on Tuesdays
  * Merging patches from Microsoft/Windows.
- Packed attributes
- Changes for MSVC
  * next-net merged
  * New driver for XSC from Yunsilicon
  * Some fixes have been merged to main to fix
GCC 15 issues.


DPDK Summit 2025 will be in Prague in May:
https://events.linuxfoundation.org/dpdk-summit/
https://www.dpdk.org/event/dpdk-summit-2025-prague/


LTS
---

Current LTS versions:

* 24.11.2
* 23.11.3
* 22.11.7

See https://core.dpdk.org/roadmap/#dates

* 20.11.10 - Will only be updated with CVE and critical fixes.
* 19.11.14 - Will only be updated with CVE and critical fixes.


* Distros
  * Debian 12 contains DPDK v22.11
  * Ubuntu 24.04 contains DPDK v23.11
  * Ubuntu 23.04 contains DPDK v22.11
  * RHEL 8/9 contains DPDK 23.11

Defects
---

* Bugzilla links, 'Bugs',  added for hosted projects
  * https://www.dpdk.org/hosted-projects/



DPDK Release Status Meetings


The DPDK Release Status Meeting is intended for DPDK Committers to discuss the
status of the master tree and sub-trees, and for project managers to track
progress or milestone dates.

The meeting occurs on every Tuesday at 14:30 DST over Jitsi on 
https://meet.jit.si/DPDK

You don't need an invite to join the meeting but if you want a calendar 
reminder just
send an email to "John McNamara john.mcnam...@intel.com" for the invite.

[PATCH 2/4] net/idpf: re-enable unused variable warnings

[Bruce Richardson]

The idpf driver was being built with warnings disabled for unused
variables. However, while the warning was being disabled in the base
code directory, all suppressed warnings were in the main directory.
Therefore, just remove the unused variables and re-enable the
warnings.

Signed-off-by: Bruce Richardson 
---
 drivers/net/intel/idpf/base/meson.build   | 9 -
 drivers/net/intel/idpf/idpf_common_rxtx.c | 2 --
 drivers/net/intel/idpf/idpf_common_virtchnl.c | 4 ++--
 3 files changed, 2 insertions(+), 13 deletions(-)

diff --git a/drivers/net/intel/idpf/base/meson.build 
b/drivers/net/intel/idpf/base/meson.build
index f30ec7dfc2..7316e0a805 100644
--- a/drivers/net/intel/idpf/base/meson.build
+++ b/drivers/net/intel/idpf/base/meson.build
@@ -5,12 +5,3 @@ sources += files(
 'idpf_controlq.c',
 'idpf_controlq_setup.c',
 )
-
-error_cflags = [
-'-Wno-unused-variable',
-]
-foreach flag: error_cflags
-if cc.has_argument(flag)
-cflags += flag
-endif
-endforeach
diff --git a/drivers/net/intel/idpf/idpf_common_rxtx.c 
b/drivers/net/intel/idpf/idpf_common_rxtx.c
index a04e54ce26..9b17279181 100644
--- a/drivers/net/intel/idpf/idpf_common_rxtx.c
+++ b/drivers/net/intel/idpf/idpf_common_rxtx.c
@@ -1178,7 +1178,6 @@ idpf_dp_singleq_recv_scatter_pkts(void *rx_queue, struct 
rte_mbuf **rx_pkts,
struct rte_mbuf *last_seg = rxq->pkt_last_seg;
struct rte_mbuf *rxm;
struct rte_mbuf *nmb;
-   struct rte_eth_dev *dev;
const uint32_t *ptype_tbl = rxq->adapter->ptype_tbl;
uint16_t rx_id = rxq->rx_tail;
uint16_t rx_packet_len;
@@ -1310,7 +1309,6 @@ idpf_xmit_cleanup(struct idpf_tx_queue *txq)
uint16_t nb_tx_desc = txq->nb_tx_desc;
uint16_t desc_to_clean_to;
uint16_t nb_tx_to_clean;
-   uint16_t i;
 
volatile struct idpf_base_tx_desc *txd = txq->tx_ring;
 
diff --git a/drivers/net/intel/idpf/idpf_common_virtchnl.c 
b/drivers/net/intel/idpf/idpf_common_virtchnl.c
index de511da788..0ae1d55d79 100644
--- a/drivers/net/intel/idpf/idpf_common_virtchnl.c
+++ b/drivers/net/intel/idpf/idpf_common_virtchnl.c
@@ -362,7 +362,7 @@ idpf_vc_queue_grps_add(struct idpf_vport *vport,
 {
struct idpf_adapter *adapter = vport->adapter;
struct idpf_cmd_info args;
-   int size, qg_info_size;
+   int size;
int err = -1;
 
size = sizeof(*p2p_queue_grps_info) +
@@ -1044,7 +1044,7 @@ int idpf_vc_rxq_config_by_info(struct idpf_vport *vport, 
struct virtchnl2_rxq_in
struct idpf_adapter *adapter = vport->adapter;
struct virtchnl2_config_rx_queues *vc_rxqs = NULL;
struct idpf_cmd_info args;
-   int size, err, i;
+   int size, err;
 
size = sizeof(*vc_rxqs) + (num_qs - 1) *
sizeof(struct virtchnl2_rxq_info);
-- 
2.43.0

[PATCH 1/4] drivers: merge common and net idpf drivers

[Bruce Richardson]

Rather than having some of the idpf code split out into the "common"
directory, used by both a net/idpf and a net/cpfl driver, we can
merge all idpf code together under net/idpf and have the cpfl driver
depend on "net/idpf" rather than "common/idpf".

Signed-off-by: Bruce Richardson 
---
 drivers/common/idpf/meson.build   | 34 ---
 drivers/common/meson.build|  1 -
 drivers/net/intel/cpfl/meson.build|  2 +-
 .../{common => net/intel}/idpf/base/README|  0
 .../intel}/idpf/base/idpf_alloc.h |  0
 .../intel}/idpf/base/idpf_controlq.c  |  0
 .../intel}/idpf/base/idpf_controlq.h  |  0
 .../intel}/idpf/base/idpf_controlq_api.h  |  0
 .../intel}/idpf/base/idpf_controlq_setup.c|  0
 .../intel}/idpf/base/idpf_devids.h|  0
 .../intel}/idpf/base/idpf_lan_pf_regs.h   |  0
 .../intel}/idpf/base/idpf_lan_txrx.h  |  0
 .../intel}/idpf/base/idpf_lan_vf_regs.h   |  0
 .../intel}/idpf/base/idpf_osdep.h |  0
 .../intel}/idpf/base/idpf_prototype.h |  0
 .../intel}/idpf/base/idpf_type.h  |  0
 .../intel}/idpf/base/meson.build  |  0
 .../intel}/idpf/base/siov_regs.h  |  0
 .../intel}/idpf/base/virtchnl2.h  |  0
 .../intel}/idpf/base/virtchnl2_lan_desc.h |  0
 .../intel}/idpf/idpf_common_device.c  |  0
 .../intel}/idpf/idpf_common_device.h  |  0
 .../intel}/idpf/idpf_common_logs.h|  0
 .../intel}/idpf/idpf_common_rxtx.c|  0
 .../intel}/idpf/idpf_common_rxtx.h|  0
 .../intel}/idpf/idpf_common_rxtx_avx512.c |  0
 .../intel}/idpf/idpf_common_virtchnl.c|  0
 .../intel}/idpf/idpf_common_virtchnl.h|  0
 drivers/net/intel/idpf/meson.build| 20 +--
 .../{common => net/intel}/idpf/version.map|  0
 drivers/net/meson.build   |  2 +-
 31 files changed, 20 insertions(+), 39 deletions(-)
 delete mode 100644 drivers/common/idpf/meson.build
 rename drivers/{common => net/intel}/idpf/base/README (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_alloc.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_controlq.c (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_controlq.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_controlq_api.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_controlq_setup.c (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_devids.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_lan_pf_regs.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_lan_txrx.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_lan_vf_regs.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_osdep.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_prototype.h (100%)
 rename drivers/{common => net/intel}/idpf/base/idpf_type.h (100%)
 rename drivers/{common => net/intel}/idpf/base/meson.build (100%)
 rename drivers/{common => net/intel}/idpf/base/siov_regs.h (100%)
 rename drivers/{common => net/intel}/idpf/base/virtchnl2.h (100%)
 rename drivers/{common => net/intel}/idpf/base/virtchnl2_lan_desc.h (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_device.c (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_device.h (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_logs.h (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_rxtx.c (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_rxtx.h (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_rxtx_avx512.c (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_virtchnl.c (100%)
 rename drivers/{common => net/intel}/idpf/idpf_common_virtchnl.h (100%)
 rename drivers/{common => net/intel}/idpf/version.map (100%)

diff --git a/drivers/common/idpf/meson.build b/drivers/common/idpf/meson.build
deleted file mode 100644
index 46fd45c03b..00
--- a/drivers/common/idpf/meson.build
+++ /dev/null
@@ -1,34 +0,0 @@
-# SPDX-License-Identifier: BSD-3-Clause
-# Copyright(c) 2022 Intel Corporation
-
-if dpdk_conf.get('RTE_IOVA_IN_MBUF') == 0
-subdir_done()
-endif
-
-includes += include_directories('../iavf')
-
-deps += ['mbuf']
-
-sources = files(
-'idpf_common_device.c',
-'idpf_common_rxtx.c',
-'idpf_common_virtchnl.c',
-)
-
-if arch_subdir == 'x86'
-if cc_has_avx512
-cflags += ['-DCC_AVX512_SUPPORT']
-avx512_args = cflags + cc_avx512_flags
-if cc.has_argument('-march=skylake-avx512')
-avx512_args += '-march=skylake-avx512'
-endif
-idpf_common_avx512_lib = static_library('idpf_common_avx512_lib',
-'idpf_common_rxtx_avx512.c',
-dependencies: [static_rte_mbuf,],
-include_directories: includes,
-c_args: avx512_args)
-objs += 
idpf_common_avx512_lib.extract_objects('idpf_c

[PATCH] ptr_compress: fix build for Arm SVE enabled SoC

[Nithin Dabilpuram]

Fix compilation issue with const conversion.

Fixes: 077596a4b077 ("ptr_compress: add pointer compression library")
Cc: sta...@dpdk.org

Signed-off-by: Nithin Dabilpuram 
---
 lib/ptr_compress/rte_ptr_compress.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/ptr_compress/rte_ptr_compress.h 
b/lib/ptr_compress/rte_ptr_compress.h
index 9742a9594a..231d96c518 100644
--- a/lib/ptr_compress/rte_ptr_compress.h
+++ b/lib/ptr_compress/rte_ptr_compress.h
@@ -134,7 +134,7 @@ rte_ptr_compress_32_shift(void *ptr_base, void * const 
*src_table,
svuint64_t v_ptr_table;
do {
svbool_t pg = svwhilelt_b64(i, n);
-   v_ptr_table = svld1_u64(pg, (uint64_t *)src_table + i);
+   v_ptr_table = svld1_u64(pg, (const uint64_t *)src_table + i);
v_ptr_table = svsub_x(pg, v_ptr_table, (uint64_t)ptr_base);
v_ptr_table = svlsr_x(pg, v_ptr_table, bit_shift);
svst1w(pg, &dest_table[i], v_ptr_table);
@@ -260,7 +260,7 @@ rte_ptr_compress_16_shift(void *ptr_base, void * const 
*src_table,
svuint64_t v_ptr_table;
do {
svbool_t pg = svwhilelt_b64(i, n);
-   v_ptr_table = svld1_u64(pg, (uint64_t *)src_table + i);
+   v_ptr_table = svld1_u64(pg, (const uint64_t *)src_table + i);
v_ptr_table = svsub_x(pg, v_ptr_table, (uint64_t)ptr_base);
v_ptr_table = svlsr_x(pg, v_ptr_table, bit_shift);
svst1h(pg, &dest_table[i], v_ptr_table);
-- 
2.34.1

[PATCH 3/4] drivers: move iavf common folder to iavf net

[Bruce Richardson]

The common/iavf driver folder contains the base code for the iavf
driver, which is also linked against by the ice driver and others.
However, there is no need for this to be in common, and we can
move it to the net/intel/iavf as a base code driver. This involves
updating dependencies that were on common/iavf to net/iavf

Signed-off-by: Bruce Richardson 
---
 drivers/common/iavf/version.map| 13 -
 drivers/common/meson.build |  1 -
 .../{common/iavf => net/intel/iavf/base}/README|  0
 .../iavf => net/intel/iavf/base}/iavf_adminq.c |  0
 .../iavf => net/intel/iavf/base}/iavf_adminq.h |  0
 .../iavf => net/intel/iavf/base}/iavf_adminq_cmd.h |  0
 .../iavf => net/intel/iavf/base}/iavf_alloc.h  |  0
 .../iavf => net/intel/iavf/base}/iavf_common.c |  0
 .../iavf => net/intel/iavf/base}/iavf_devids.h |  0
 .../iavf => net/intel/iavf/base}/iavf_impl.c   |  0
 .../iavf => net/intel/iavf/base}/iavf_osdep.h  |  0
 .../iavf => net/intel/iavf/base}/iavf_prototype.h  |  0
 .../iavf => net/intel/iavf/base}/iavf_register.h   |  0
 .../iavf => net/intel/iavf/base}/iavf_status.h |  0
 .../iavf => net/intel/iavf/base}/iavf_type.h   |  0
 .../iavf => net/intel/iavf/base}/meson.build   |  0
 .../iavf => net/intel/iavf/base}/virtchnl.h|  0
 .../intel/iavf/base}/virtchnl_inline_ipsec.h   |  0
 drivers/net/intel/iavf/meson.build | 13 +
 drivers/net/intel/iavf/version.map | 14 ++
 drivers/net/intel/ice/meson.build  |  7 +++
 drivers/net/intel/idpf/meson.build |  2 +-
 22 files changed, 27 insertions(+), 23 deletions(-)
 delete mode 100644 drivers/common/iavf/version.map
 rename drivers/{common/iavf => net/intel/iavf/base}/README (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_adminq.c (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_adminq.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_adminq_cmd.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_alloc.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_common.c (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_devids.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_impl.c (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_osdep.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_prototype.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_register.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_status.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_type.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/meson.build (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/virtchnl.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/virtchnl_inline_ipsec.h 
(100%)

diff --git a/drivers/common/iavf/version.map b/drivers/common/iavf/version.map
deleted file mode 100644
index 6c1427cca4..00
--- a/drivers/common/iavf/version.map
+++ /dev/null
@@ -1,13 +0,0 @@
-INTERNAL {
-   global:
-
-   iavf_aq_send_msg_to_pf;
-   iavf_clean_arq_element;
-   iavf_init_adminq;
-   iavf_set_mac_type;
-   iavf_shutdown_adminq;
-   iavf_vf_parse_hw_config;
-   iavf_vf_reset;
-
-   local: *;
-};
diff --git a/drivers/common/meson.build b/drivers/common/meson.build
index e1e3149d8f..dc096aab0a 100644
--- a/drivers/common/meson.build
+++ b/drivers/common/meson.build
@@ -5,7 +5,6 @@ std_deps = ['eal']
 drivers = [
 'cpt',
 'dpaax',
-'iavf',
 'ionic',
 'mvep',
 'octeontx',
diff --git a/drivers/common/iavf/README b/drivers/net/intel/iavf/base/README
similarity index 100%
rename from drivers/common/iavf/README
rename to drivers/net/intel/iavf/base/README
diff --git a/drivers/common/iavf/iavf_adminq.c 
b/drivers/net/intel/iavf/base/iavf_adminq.c
similarity index 100%
rename from drivers/common/iavf/iavf_adminq.c
rename to drivers/net/intel/iavf/base/iavf_adminq.c
diff --git a/drivers/common/iavf/iavf_adminq.h 
b/drivers/net/intel/iavf/base/iavf_adminq.h
similarity index 100%
rename from drivers/common/iavf/iavf_adminq.h
rename to drivers/net/intel/iavf/base/iavf_adminq.h
diff --git a/drivers/common/iavf/iavf_adminq_cmd.h 
b/drivers/net/intel/iavf/base/iavf_adminq_cmd.h
similarity index 100%
rename from drivers/common/iavf/iavf_adminq_cmd.h
rename to drivers/net/intel/iavf/base/iavf_adminq_cmd.h
diff --git a/drivers/common/iavf/iavf_alloc.h 
b/drivers/net/intel/iavf/base/iavf_alloc.h
similarity index 100%
rename from drivers/common/iavf/iavf_alloc.h
rename to drivers/net/intel/iavf/base/iavf_alloc.h
diff --git a/drivers/common/iavf/iavf_common.c 
b/drivers/net/intel/iavf/base/iavf_common.c
similarity index 100%
rename from drivers/common/iavf/iavf_common.c
rename to drivers/net/intel/iavf/base/iavf_common

[PATCH] net/bnxt: remove unnecessary checks for null pointer

[Stephen Hemminger]

The function rte_free() handles a NULL pointer as no-op.
This patch was automatically generated from nullfree.cocci script.

Signed-off-by: Stephen Hemminger 
---
 drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c | 6 ++
 drivers/net/bnxt/tf_ulp/ulp_mapper.c| 3 +--
 drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c| 6 ++
 3 files changed, 5 insertions(+), 10 deletions(-)

diff --git a/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c 
b/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c
index 1770069295..f88299bbf7 100644
--- a/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c
+++ b/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c
@@ -1668,10 +1668,8 @@ int tfc_tbl_scope_cpm_alloc(struct tfc *tfcp, uint8_t 
tsid,
return 0;
 
  cleanup:
-   if (cmm_lkup != NULL)
-   rte_free(cmm_lkup);
-   if (cmm_act != NULL)
-   rte_free(cmm_act);
+   rte_free(cmm_lkup);
+   rte_free(cmm_act);
 
return rc;
 }
diff --git a/drivers/net/bnxt/tf_ulp/ulp_mapper.c 
b/drivers/net/bnxt/tf_ulp/ulp_mapper.c
index 2429ac2f1a..49cd6620dd 100644
--- a/drivers/net/bnxt/tf_ulp/ulp_mapper.c
+++ b/drivers/net/bnxt/tf_ulp/ulp_mapper.c
@@ -1359,8 +1359,7 @@ ulp_mapper_key_recipe_tbl_deinit(struct 
bnxt_ulp_mapper_data *mdata)
recipes = mdata->key_recipe_info.recipes[dir][ftype];
for (idx = 0; idx < mdata->key_recipe_info.num_recipes;
  idx++) {
-   if (recipes[idx])
-   rte_free(recipes[idx]);
+   rte_free(recipes[idx]);
}
rte_free(mdata->key_recipe_info.recipes[dir][ftype]);
mdata->key_recipe_info.recipes[dir][ftype] = NULL;
diff --git a/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c 
b/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c
index 5fa8e240db..85b72f328c 100644
--- a/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c
+++ b/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c
@@ -158,11 +158,9 @@ ulp_sc_mgr_deinit(struct bnxt_ulp_context *ctxt)
if (!ulp_sc_info)
return -EINVAL;
 
-   if (ulp_sc_info->stats_cache_tbl)
-   rte_free(ulp_sc_info->stats_cache_tbl);
+   rte_free(ulp_sc_info->stats_cache_tbl);
 
-   if (ulp_sc_info->read_data)
-   rte_free(ulp_sc_info->read_data);
+   rte_free(ulp_sc_info->read_data);
 
rte_free(ulp_sc_info);
 
-- 
2.45.2

[PATCH 0/4] remove common iavf and idpf drivers

[Bruce Richardson]

The iavf and idpf common directories were used only to share code
between multiple net drivers and did not need to be drivers in their own
right, since it is just as easy to have a dependency from one net driver
on another as a net driver on a common one.

This patchset therefore aims to eliminate the two unnecessary common
drivers. It does so as follows:

* merging common/idpf into net/idpf and updating the cpfl dependency to
  point to the net driver.
* merging common/iavf into net/iavf and similarly updating the
  dependencies, including the paths from idpf (which does not directly
  depend on iavf, but does make use of the definitions in the iavf
  header files).

Separately, two other cleanups are done - one to remove an unnecessary
warning disable flag. The second is a little more complex - it makes the
dependency between ice and iavf an optional one, by having ice compile
in the necessary iavf shared code files in case iavf is disabled in the
build.

Bruce Richardson (4):
  drivers: merge common and net idpf drivers
  net/idpf: re-enable unused variable warnings
  drivers: move iavf common folder to iavf net
  net/intel: allow building ice driver without iavf

 drivers/common/iavf/version.map   | 13 ---
 drivers/common/idpf/meson.build   | 34 ---
 drivers/common/meson.build|  2 --
 drivers/net/intel/cpfl/meson.build|  2 +-
 .../iavf => net/intel/iavf/base}/README   |  0
 .../intel/iavf/base}/iavf_adminq.c|  0
 .../intel/iavf/base}/iavf_adminq.h|  0
 .../intel/iavf/base}/iavf_adminq_cmd.h|  0
 .../iavf => net/intel/iavf/base}/iavf_alloc.h |  0
 .../intel/iavf/base}/iavf_common.c|  0
 .../intel/iavf/base}/iavf_devids.h|  0
 .../iavf => net/intel/iavf/base}/iavf_impl.c  |  0
 .../iavf => net/intel/iavf/base}/iavf_osdep.h |  0
 .../intel/iavf/base}/iavf_prototype.h |  8 +
 .../intel/iavf/base}/iavf_register.h  |  0
 .../intel/iavf/base}/iavf_status.h|  0
 .../iavf => net/intel/iavf/base}/iavf_type.h  |  0
 .../iavf => net/intel/iavf/base}/meson.build  |  0
 .../iavf => net/intel/iavf/base}/virtchnl.h   |  0
 .../intel/iavf/base}/virtchnl_inline_ipsec.h  |  0
 drivers/net/intel/iavf/meson.build| 13 ---
 drivers/net/intel/iavf/version.map| 14 
 drivers/net/intel/ice/meson.build | 18 +++---
 .../{common => net/intel}/idpf/base/README|  0
 .../intel}/idpf/base/idpf_alloc.h |  0
 .../intel}/idpf/base/idpf_controlq.c  |  0
 .../intel}/idpf/base/idpf_controlq.h  |  0
 .../intel}/idpf/base/idpf_controlq_api.h  |  0
 .../intel}/idpf/base/idpf_controlq_setup.c|  0
 .../intel}/idpf/base/idpf_devids.h|  0
 .../intel}/idpf/base/idpf_lan_pf_regs.h   |  0
 .../intel}/idpf/base/idpf_lan_txrx.h  |  0
 .../intel}/idpf/base/idpf_lan_vf_regs.h   |  0
 .../intel}/idpf/base/idpf_osdep.h |  0
 .../intel}/idpf/base/idpf_prototype.h |  0
 .../intel}/idpf/base/idpf_type.h  |  0
 .../intel}/idpf/base/meson.build  |  9 -
 .../intel}/idpf/base/siov_regs.h  |  0
 .../intel}/idpf/base/virtchnl2.h  |  0
 .../intel}/idpf/base/virtchnl2_lan_desc.h |  0
 .../intel}/idpf/idpf_common_device.c  |  0
 .../intel}/idpf/idpf_common_device.h  |  0
 .../intel}/idpf/idpf_common_logs.h|  0
 .../intel}/idpf/idpf_common_rxtx.c|  2 --
 .../intel}/idpf/idpf_common_rxtx.h|  0
 .../intel}/idpf/idpf_common_rxtx_avx512.c |  0
 .../intel}/idpf/idpf_common_virtchnl.c|  4 +--
 .../intel}/idpf/idpf_common_virtchnl.h|  0
 drivers/net/intel/idpf/meson.build| 20 +--
 .../{common => net/intel}/idpf/version.map|  0
 drivers/net/meson.build   |  2 +-
 51 files changed, 67 insertions(+), 74 deletions(-)
 delete mode 100644 drivers/common/iavf/version.map
 delete mode 100644 drivers/common/idpf/meson.build
 rename drivers/{common/iavf => net/intel/iavf/base}/README (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_adminq.c (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_adminq.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_adminq_cmd.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_alloc.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_common.c (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_devids.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_impl.c (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_osdep.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_prototype.h (94%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_register.h (100%)
 rename drivers/{common/iavf => net/intel/iavf/base}/iavf_status.h (100%)
 rename drivers/{common

[PATCH 4/4] net/intel: allow building ice driver without iavf

[Bruce Richardson]

The ice PMD relies on a number of functions from the iavf base code,
which can be got by linking against that iavf driver. However, since
only three C files are necessary here, we can allow ice to be built
independently of iavf by including the base files directly in cases
where iavf is not part of the build. If it is part of the build, the
dependency remains as now.

Signed-off-by: Bruce Richardson 
---
 drivers/net/intel/iavf/base/iavf_prototype.h |  8 
 drivers/net/intel/ice/meson.build| 13 -
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/drivers/net/intel/iavf/base/iavf_prototype.h 
b/drivers/net/intel/iavf/base/iavf_prototype.h
index 7c43a817bb..5d2ee0a785 100644
--- a/drivers/net/intel/iavf/base/iavf_prototype.h
+++ b/drivers/net/intel/iavf/base/iavf_prototype.h
@@ -11,6 +11,14 @@
 
 #include 
 
+/* functions only need exporting if this is being built into
+ * iavf driver itself. If included in ice driver, then no export
+ */
+#ifndef RTE_NET_IAVF
+#undef __rte_internal
+#define __rte_internal
+#endif
+
 /* Prototypes for shared code functions that are not in
  * the standard function pointer structures.  These are
  * mostly because they are needed even before the init
diff --git a/drivers/net/intel/ice/meson.build 
b/drivers/net/intel/ice/meson.build
index 5faf887386..f990e62250 100644
--- a/drivers/net/intel/ice/meson.build
+++ b/drivers/net/intel/ice/meson.build
@@ -18,9 +18,20 @@ sources = files(
 
 testpmd_sources = files('ice_testpmd.c')
 
-deps += ['hash', 'net', 'net_iavf']
+deps += ['hash', 'net']
 includes += include_directories('base')
 
+if dpdk_conf.has('RTE_NET_IAVF')
+deps += 'net_iavf'
+else
+includes += include_directories('../iavf/base')
+sources += files(
+'../iavf/base/iavf_adminq.c',
+'../iavf/base/iavf_common.c',
+'../iavf/base/iavf_impl.c',
+)
+endif
+
 if arch_subdir == 'x86'
 sources += files('ice_rxtx_vec_sse.c')
 
-- 
2.43.0

Re: [PATCH] net/bnxt: remove unnecessary checks for null pointer

[Ajit Khaparde]

On Tue, Jan 28, 2025 at 9:42 AM Stephen Hemminger
 wrote:
>
> The function rte_free() handles a NULL pointer as no-op.
> This patch was automatically generated from nullfree.cocci script.
>
> Signed-off-by: Stephen Hemminger 
Acked-by: Ajit Khaparde 

> ---
>  drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c | 6 ++
>  drivers/net/bnxt/tf_ulp/ulp_mapper.c| 3 +--
>  drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c| 6 ++
>  3 files changed, 5 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c 
> b/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c
> index 1770069295..f88299bbf7 100644
> --- a/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c
> +++ b/drivers/net/bnxt/tf_core/v3/tfc_tbl_scope.c
> @@ -1668,10 +1668,8 @@ int tfc_tbl_scope_cpm_alloc(struct tfc *tfcp, uint8_t 
> tsid,
> return 0;
>
>   cleanup:
> -   if (cmm_lkup != NULL)
> -   rte_free(cmm_lkup);
> -   if (cmm_act != NULL)
> -   rte_free(cmm_act);
> +   rte_free(cmm_lkup);
> +   rte_free(cmm_act);
>
> return rc;
>  }
> diff --git a/drivers/net/bnxt/tf_ulp/ulp_mapper.c 
> b/drivers/net/bnxt/tf_ulp/ulp_mapper.c
> index 2429ac2f1a..49cd6620dd 100644
> --- a/drivers/net/bnxt/tf_ulp/ulp_mapper.c
> +++ b/drivers/net/bnxt/tf_ulp/ulp_mapper.c
> @@ -1359,8 +1359,7 @@ ulp_mapper_key_recipe_tbl_deinit(struct 
> bnxt_ulp_mapper_data *mdata)
> recipes = mdata->key_recipe_info.recipes[dir][ftype];
> for (idx = 0; idx < 
> mdata->key_recipe_info.num_recipes;
>   idx++) {
> -   if (recipes[idx])
> -   rte_free(recipes[idx]);
> +   rte_free(recipes[idx]);
> }
> rte_free(mdata->key_recipe_info.recipes[dir][ftype]);
> mdata->key_recipe_info.recipes[dir][ftype] = NULL;
> diff --git a/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c 
> b/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c
> index 5fa8e240db..85b72f328c 100644
> --- a/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c
> +++ b/drivers/net/bnxt/tf_ulp/ulp_sc_mgr.c
> @@ -158,11 +158,9 @@ ulp_sc_mgr_deinit(struct bnxt_ulp_context *ctxt)
> if (!ulp_sc_info)
> return -EINVAL;
>
> -   if (ulp_sc_info->stats_cache_tbl)
> -   rte_free(ulp_sc_info->stats_cache_tbl);
> +   rte_free(ulp_sc_info->stats_cache_tbl);
>
> -   if (ulp_sc_info->read_data)
> -   rte_free(ulp_sc_info->read_data);
> +   rte_free(ulp_sc_info->read_data);
>
> rte_free(ulp_sc_info);
>
> --
> 2.45.2
>


smime.p7s
Description: S/MIME Cryptographic Signature

Re: [PATCH v9 00/15] net/zxdh: updated net zxdh driver

[Stephen Hemminger]

On Tue, 21 Jan 2025 11:44:18 +0800
Junlong Wang  wrote:

> V9:
>   - Remove unnecessary __rte_packed* in stats structure.
> 
> V8:
>   - using __rte_packed_begin/__rte_packed_end replace __rte_packed.
> 
> V7:
>   - resolved warning '-Waddress-of-packed-member'
> in function 'zxdh_dev_rss_reta_update'.
> 
> V6:
>   - Remove unnecessary __rte_packed in the virtqueue structure and others.
>   - Remove Some blank before or after log message,
> and remove some end with period in log message.
> 
> V5:
>   - Simplify the notify_data part in the zxdh_notify_queue function.
>   - Replace rte_zmalloc with rte_calloc in the rss_reta_update function.
>   - Remove unnecessary check in mtu_set function.
> 
> V4:
>   - resolved ci compile issues.
> 
> V3:
>   - use rte_zmalloc and rte_calloc to avoid memset.
>   - remove unnecessary initialization, which first usage will set.
>   - adjust some function which is always return 0, changed to void 
> and skip the ASSERTION later.
>   - resolved some WARNING:MACRO_ARG_UNUSED issues.
>   - resolved some other issues.
> 
> V2:
>   - resolve code style and github-robot build issue.
> 
> V1:
>   - updated net zxdh driver
> provided insert/delete/get table code funcs.
> provided link/mac/vlan/promiscuous/rss/mtu ops.
> 
> Junlong Wang (15):
>   net/zxdh: zxdh np init implementation
>   net/zxdh: zxdh np uninit implementation
>   net/zxdh: port tables init implementations
>   net/zxdh: port tables unint implementations
>   net/zxdh: rx/tx queue setup and intr enable
>   net/zxdh: dev start/stop ops implementations
>   net/zxdh: provided dev simple tx implementations
>   net/zxdh: provided dev simple rx implementations
>   net/zxdh: link info update, set link up/down
>   net/zxdh: mac set/add/remove ops implementations
>   net/zxdh: promisc/allmulti ops implementations
>   net/zxdh: vlan filter/ offload ops implementations
>   net/zxdh: rss hash config/update, reta update/get
>   net/zxdh: basic stats ops implementations
>   net/zxdh: mtu update ops implementations
> 
>  doc/guides/nics/features/zxdh.ini  |   18 +
>  doc/guides/nics/zxdh.rst   |   17 +
>  drivers/net/zxdh/meson.build   |4 +
>  drivers/net/zxdh/zxdh_common.c |   28 +-
>  drivers/net/zxdh/zxdh_common.h |1 +
>  drivers/net/zxdh/zxdh_ethdev.c |  602 +++-
>  drivers/net/zxdh/zxdh_ethdev.h |   48 +-
>  drivers/net/zxdh/zxdh_ethdev_ops.c | 1573 +
>  drivers/net/zxdh/zxdh_ethdev_ops.h |   80 ++
>  drivers/net/zxdh/zxdh_msg.c|  205 ++-
>  drivers/net/zxdh/zxdh_msg.h|  232 
>  drivers/net/zxdh/zxdh_np.c | 2060 
>  drivers/net/zxdh/zxdh_np.h |  579 
>  drivers/net/zxdh/zxdh_pci.c|   27 +-
>  drivers/net/zxdh/zxdh_pci.h|9 +-
>  drivers/net/zxdh/zxdh_queue.c  |  242 +++-
>  drivers/net/zxdh/zxdh_queue.h  |  189 ++-
>  drivers/net/zxdh/zxdh_rxtx.c   |  804 +++
>  drivers/net/zxdh/zxdh_rxtx.h   |   23 +-
>  drivers/net/zxdh/zxdh_tables.c |  794 +++
>  drivers/net/zxdh/zxdh_tables.h |  231 
>  21 files changed, 7671 insertions(+), 95 deletions(-)
>  create mode 100644 drivers/net/zxdh/zxdh_ethdev_ops.c
>  create mode 100644 drivers/net/zxdh/zxdh_ethdev_ops.h
>  create mode 100644 drivers/net/zxdh/zxdh_np.c
>  create mode 100644 drivers/net/zxdh/zxdh_np.h
>  create mode 100644 drivers/net/zxdh/zxdh_rxtx.c
>  create mode 100644 drivers/net/zxdh/zxdh_tables.c
>  create mode 100644 drivers/net/zxdh/zxdh_tables.h
> 


Did minor conflict resolution on release notes and
applied to dpdk-next-net (for-main).

Re: [PATCH 4/4] net/netvsc: cache device parameters for hot plug events

[Stephen Hemminger]

On Mon, 27 Jan 2025 17:35:06 -0800
lon...@linuxonhyperv.com wrote:

> @@ -1409,9 +1424,6 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
>   ret_stop = hn_dev_stop(eth_dev);
>   hn_dev_close(eth_dev);
>  
> - free(hv->vf_devargs);
> - hv->vf_devargs = NULL;
> -
>   hn_detach(hv);
>   hn_chim_uninit(eth_dev);
>   rte_vmbus_chan_close(hv->channels[0]);
> @@ -1423,6 +1435,61 @@ eth_hn_dev_uninit(struct rte_eth_dev *eth_dev)
>   return ret_stop;
>  }
>  
> +static int populate_cache_list(void)
> +{
> + int ret;
> + struct rte_devargs *da;
> +
> + rte_spinlock_lock(&netvsc_lock);
> + da_cache_usage++;
> + if (da_cache_usage > 1) {
> + ret = 0;
> + goto out;
> + }
> +
> + LIST_INIT(&da_cache_list);
> + RTE_EAL_DEVARGS_FOREACH("pci", da) {
> + struct da_cache *cache;
> +
> + cache = rte_zmalloc("NETVSC-HOTADD", sizeof(*cache), 
> rte_mem_page_size());
> + if (!cache) {
> + ret = -ENOMEM;
> + goto out;
> + }
> +
> + strncpy(cache->name, da->name, sizeof(da->name));
> + cache->drv_str = strdup(da->drv_str);
> + if (!cache->drv_str) {
> + rte_free(cache);
> + ret = -ENOMEM;
> + goto out;
> + }
> +
> + LIST_INSERT_HEAD(&da_cache_list, cache, list);
> + }

Why do you need to cache entry to be page aligned, that seems unnecessary 
wasteful?
Why does it need to be huge pages? versus normal malloc?
The string is coming from malloc (strdup) so it can't be used by secondary 
process.

Since you are allocating a devargs cache entry, you could just as well use
flexible array where entry was like:
struct da_cache {
LIST_ENTRY(da_cache) list;
char name[RTE_DEV_NAME_MAX_LEN];
char drv_str[];
};


cache = malloc(sizeof(*cache) + strlen(da->drv_str) + 1);
...
strcpy(cache->drv_str, da->drv_str);

Re: [PATCH 4/4] net/netvsc: cache device parameters for hot plug events

[Stephen Hemminger]

On Mon, 27 Jan 2025 17:35:06 -0800
lon...@linuxonhyperv.com wrote:

> From: Long Li 
> 
> If a device is hot removed and hot plugged, it needs the same driver
> parameters that are passed to EAL. However, during device removal, all
> EAL driver parameters are freed as part of the cleanup.
> 
> Cache those driver parameters for future hot plug events. Because we don't
> know which device will show up, cache all the PCI driver parameters.
> 
> Signed-off-by: Long Li 

*Build Failed #1:
OS: OpenAnolis8.9-64
Target: x86_64-native-linuxapp-gcc
FAILED: drivers/libtmp_rte_net_netvsc.a.p/net_netvsc_hn_ethdev.c.o 
gcc -Idrivers/libtmp_rte_net_netvsc.a.p -Idrivers -I../drivers 
-Idrivers/net/netvsc -I../drivers/net/netvsc -Ilib/ethdev -I../lib/ethdev -I. 
-I.. -Iconfig -I../config -Ilib/eal/include -I../lib/eal/include 
-Ilib/eal/linux/include -I../lib/eal/linux/include -Ilib/eal/x86/include 
-I../lib/eal/x86/include -I../kernel/linux -Ilib/eal/common -I../lib/eal/common 
-Ilib/eal -I../lib/eal -Ilib/kvargs -I../lib/kvargs -Ilib/log -I../lib/log 
-Ilib/metrics -I../lib/metrics -Ilib/telemetry -I../lib/telemetry -Ilib/net 
-I../lib/net -Ilib/mbuf -I../lib/mbuf -Ilib/mempool -I../lib/mempool -Ilib/ring 
-I../lib/ring -Ilib/meter -I../lib/meter -Idrivers/bus/pci -I../drivers/bus/pci 
-I../drivers/bus/pci/linux -Ilib/pci -I../lib/pci -Idrivers/bus/vdev 
-I../drivers/bus/vdev -Idrivers/bus/vmbus -I../drivers/bus/vmbus 
-I../drivers/bus/vmbus/linux -fdiagnostics-color=always -D_FILE_OFFSET_BITS=64 
-Wall -Winvalid-pch -Wextra -Werror -std=c11 -O3 -include rte_config.h 
-Wcast-qual -Wdeprecated -Wformat -Wformat-nonliteral -Wformat-security 
-Wmissing-declarations -Wmissing-prototypes -Wnested-externs 
-Wold-style-definition -Wpointer-arith -Wsign-compare -Wstrict-prototypes 
-Wundef -Wwrite-strings -Wno-packed-not-aligned -Wno-missing-field-initializers 
-D_GNU_SOURCE -fPIC -march=native -mrtm -DALLOW_EXPERIMENTAL_API 
-DALLOW_INTERNAL_API -Wno-format-truncation 
-DRTE_LOG_DEFAULT_LOGTYPE=pmd.net.netvsc -MD -MQ 
drivers/libtmp_rte_net_netvsc.a.p/net_netvsc_hn_ethdev.c.o -MF 
drivers/libtmp_rte_net_netvsc.a.p/net_netvsc_hn_ethdev.c.o.d -o 
drivers/libtmp_rte_net_netvsc.a.p/net_netvsc_hn_ethdev.c.o -c 
../drivers/net/netvsc/hn_ethdev.c
../drivers/net/netvsc/hn_ethdev.c: In function ‘populate_cache_list’:
../drivers/net/netvsc/hn_ethdev.c:1460:40: error: argument to ‘sizeof’ in 
‘strncpy’ call is the same expression as the source; did you mean to use the 
size of the destination? [-Werror=sizeof-pointer-memaccess]
   strncpy(cache->name, da->name, sizeof(da->name));
^
../drivers/net/netvsc/hn_ethdev.c: In function ‘eth_hn_probe’:
../drivers/net/netvsc/hn_ethdev.c:1502:5: error: ‘ret’ may be used 
uninitialized in this function [-Werror=maybe-uninitialized]
  if (ret)
 ^

Re: [PATCH v3 0/6] queue start/stop and deferred checks

[Stephen Hemminger]

On Mon, 16 Dec 2024 11:11:32 -0800
Stephen Hemminger  wrote:

> Recent zxdh driver review raised the question of why should
> drivers have to check rx_conf for deferred start support.
> This can be better handled across all drivers at ethdev level.
> 
> Also found some drivers were incorrect in feature flags about
> handling of queue start/stop.
> 
> v3 - fix doc about queue start/stop for some drivers
> 
> Stephen Hemminger (6):
>   ethdev: check that device supports deferred start
>   doc: fix feature flags for queue start/stop
>   net/dpaa: remove unnecessary deferred start check
>   net/dpaa2: remove unneeded deferred start check
>   net/enetfec: remove unneeded deferred start check
>   net/virtio: remove unneeded deferred start check

I am deferring this patchset for now, need ACK from some of the drivers 
impacted first

RE: [RFC PATCH] eventdev: adapter API to configure multiple Rx queues

[Naga Harish K, S V]

> -Original Message-
> From: Shijith Thotton 
> Sent: Friday, January 24, 2025 3:30 PM
> To: Naga Harish K, S V ; dev@dpdk.org
> Cc: Pavan Nikhilesh Bhagavatula ; Pathak,
> Pravin ; Hemant Agrawal
> ; Sachin Saxena ;
> Mattias R_nnblom ; Jerin Jacob
> ; Liang Ma ; Mccarthy, Peter
> ; Van Haaren, Harry
> ; Carrillo, Erik G ;
> Gujjar, Abhinandan S ; Amit Prakash Shukla
> ; Burakov, Anatoly
> 
> Subject: RE: [RFC PATCH] eventdev: adapter API to configure multiple Rx
> queues
> 
> >> >> >> >> >>> This RFC introduces a new API,
> >> >> >> >> >>> rte_event_eth_rx_adapter_queues_add(),
> >> >> >> >> >>> designed to enhance the flexibility of configuring
> >> >> >> >> >>> multiple Rx queues in eventdev Rx adapter.
> >> >> >> >> >>>
> >> >> >> >> >>> The existing rte_event_eth_rx_adapter_queue_add() API
> >> >> >> >> >>> supports adding multiple queues by specifying
> >> >> >> >> >>> rx_queue_id = -1, but it lacks the ability to
> >> >> >> >> >apply
> >> >> >> >> >>> specific configurations to each of the added queues.
> >> >> >> >> >>>
> >> >> >> >> >>
> >> >> >> >> >>The application can still use the existing
> >> >> >> >> >>rte_event_eth_rx_adapter_queue_add() API in a loop with
> >> >> >> >> >>different configurations for different queues.
> >> >> >> >> >>
> >> >> >> >> >>The proposed API is not enabling new features that cannot
> >> >> >> >> >>be achieved with the existing API.
> >> >> >> >> >>Adding new APIs without much usefulness causes unnecessary
> >> >> >> >> >>complexity/confusion for users.
> >> >> >> >> >>
> >> >>
> >> >> The eth_rx_adapter_queue_add eventdev PMD operation can be
> updated
> >> to
> >> >> support burst mode. Internally, both the new and existing APIs can
> >> >> utilize this updated operation. This enables applications to use
> >> >> either API and achieve
> >> >the
> >> >> same results while adding a single queue. For adding multiple RX
> >> >> queues to
> >> >the
> >> >> adapter, the new API can be used as it is not supported by the old API.
> >> >>
> >> >
> >> >Not all platforms implement the eventdev PMD operation for
> >> >eth_rx_adapter_queue_add, so this does not apply to all platforms.
> >> >
> >>
> >> Yes, but there are hardware PMDs that implement
> >eth_rx_adapter_queue_add
> >> op, and I am looking for a solution that works for both cases.
> >>
> >> The idea is to use the new eventdev PMD operation
> >> (eth_rx_adapter_queues_add) within the
> >> rte_event_eth_rx_adapter_queue_add() API. The parameters of this API
> >> can be easily mapped to and supported by the new PMD operation.
> >>
> >
> >This requires a change to the rte_event_eth_rx_adapter_queue_add()
> >stable API parameters.
> >This is an ABI breakage and may not be possible now.
> >It requires changes to many current applications that are using the
> >rte_event_eth_rx_adapter_queue_add() stable API.
> >
> 
> What I meant by mapping was to retain the stable API parameters as they are.
> Internally, the API can use the proposed eventdev PMD operation
> (eth_rx_adapter_queues_add) without causing an ABI break, as shown below.
> 
> int rte_event_eth_rx_adapter_queue_add(uint8_t id, uint16_t eth_dev_id,
> int32_t rx_queue_id,
> const struct rte_event_eth_rx_adapter_queue_conf *conf) {
> if (rx_queue_id == -1)
> dev->dev_ops->eth_rx_adapter_queues_add)(
> dev, &rte_eth_devices[eth_dev_id], 0,
> conf, 0);
> else
> dev->dev_ops->eth_rx_adapter_queues_add)(
> dev, &rte_eth_devices[eth_dev_id], &rx_queue_id,
> conf, 1);
> }
> 
> With above change, old op (eth_rx_adapter_queue_add) can be removed as
> both API (stable and proposed) will be using eth_rx_adapter_queues_add.
> 

The whole idea is not to have the proposed API, as it does not add any new 
feature but just a combination of existing API.
It is already discussed in the previous threads.
The internal implementation details are not a concern.

> >> typedef int (*eventdev_eth_rx_adapter_queues_add_t)(
> >> const struct rte_eventdev *dev,
> >> const struct rte_eth_dev *eth_dev,
> >> int32_t rx_queue_id[],
> >> const struct rte_event_eth_rx_adapter_queue_conf queue_conf[],
> >> uint16_t nb_rx_queues);
> >>
> >> With this, the old PMD op (eth_rx_adapter_queue_add) can be removed.
> >>
> >> >> >> >> >
> >> >> >> >> >The new API was introduced because the existing API does
> >> >> >> >> >not support adding multiple queues with specific configurations.
> >> >> >> >> >It serves as a burst variant of the existing API, like many
> >> >> >> >> >other APIs in
> >> >> DPDK.
> >> >> >> >> >
> >> >> >> >
> >> >> >> >The other burst APIs may be there for dataplane
> >> >> >> >functionalities, but may not be for the control plane 
> >> >> >> >functionalities.
> >> >> >> >
> >> >> >>
> >> >> >> rte_acl_add_rules() is an example of burst API in control path.
> >> >> >>
> >> >> >
> 

RE: [RFC PATCH] eventdev: adapter API to configure multiple Rx queues

[Jerin Jacob]

> -Original Message-
> From: Naga Harish K, S V 
> Sent: Wednesday, January 29, 2025 10:35 AM
> To: Shijith Thotton ; dev@dpdk.org
> Cc: Pavan Nikhilesh Bhagavatula ; Pathak, Pravin
> ; Hemant Agrawal ;
> Sachin Saxena ; Mattias R_nnblom
> ; Jerin Jacob ; Liang
> Ma ; Mccarthy, Peter ;
> Van Haaren, Harry ; Carrillo, Erik G
> ; Gujjar, Abhinandan S
> ; Amit Prakash Shukla
> ; Burakov, Anatoly 
> Subject: [EXTERNAL] RE: [RFC PATCH] eventdev: adapter API to configure
> multiple Rx queues
> > >
> > >This requires a change to the rte_event_eth_rx_adapter_queue_add()
> > >stable API parameters.
> > >This is an ABI breakage and may not be possible now.
> > >It requires changes to many current applications that are using the
> > >rte_event_eth_rx_adapter_queue_add() stable API.
> > >
> >
> > What I meant by mapping was to retain the stable API parameters as they are.
> > Internally, the API can use the proposed eventdev PMD operation
> > (eth_rx_adapter_queues_add) without causing an ABI break, as shown below.
> >
> > int rte_event_eth_rx_adapter_queue_add(uint8_t id, uint16_t eth_dev_id,
> > int32_t rx_queue_id,
> > const struct rte_event_eth_rx_adapter_queue_conf *conf) {
> > if (rx_queue_id == -1)
> > dev->dev_ops->eth_rx_adapter_queues_add)(
> > dev, &rte_eth_devices[eth_dev_id], 0,
> > conf, 0);
> > else
> > dev->dev_ops->eth_rx_adapter_queues_add)(
> > dev, &rte_eth_devices[eth_dev_id], &rx_queue_id,
> > conf, 1);
> > }
> >
> > With above change, old op (eth_rx_adapter_queue_add) can be removed as
> > both API (stable and proposed) will be using eth_rx_adapter_queues_add.


Since this thread is not converging and looks like it is due to confusion.
I am trying to summarize my understanding to define the next steps(like if 
needed, we need to reach tech board if there are no consensus)


Problem statement:
==
1) Implementation of rte_event_eth_rx_adapter_queue_add() in HW typically uses 
an administrative
function to enable it. Typically, it translated to sending a mailbox to PF 
driver etc.
So, this function takes "time" to complete in HW implementations.
2) For SW implementations, this won't take time as there is no other actors 
involved.
3) There are customer use cases, they add 300+ 
rte_event_eth_rx_adapter_queue_add() on 
application bootup, that is introducing significant boot time for the 
application.
Number of queues are function of number of ethdev ports, number  of ethdev Rx 
queues per port
and number of event queues.


Expected outcome of problem statement:
==
1) The cases where application knows queue mapping(typically at boot time case),
application can call burst variant of rte_event_eth_rx_adapter_queue_add() 
function
to amortize the cost. Similar scheme used DPDK in control path API where 
latency is critical,
like rte_acl_add_rules() or rte_flow via template scheme.
2) Solution should not break ABI or any impact to SW drivers.
3) Avoid duplicating the code as much as possible


Proposed solution:
==
1) Update eventdev_eth_rx_adapter_queue_add_t() PMD (Internal ABI) API to take 
burst parameters
2) Add new rte_event_eth_rx_adapter_queue*s*_add() function and wire to use 
updated PMD API
3) Use rte_event_eth_rx_adapter_queue_add() as 
rte_event_eth_rx_adapter_queue*s*_add(, 1)

If so, I am not sure what is the cons of this approach, it will let to have 
optimized applications when
a) Application knows the queue mapping at priorly (typically in boot time)
b) Allow HW drivers to optimize without breaking anything for SW drivers
c) Provide applications to decide burst vs non burst selection based on the 
needed and performance requirements