From: David Marchand <david.marchand@redhat.com>
To: Stephen Hemminger <stephen@networkplumber.org>,
Tyler Retzlaff <roretzla@linux.microsoft.com>
Cc: dev@dpdk.org, Yipeng Wang <yipeng1.wang@intel.com>,
Sameh Gobriel <sameh.gobriel@intel.com>,
Bruce Richardson <bruce.richardson@intel.com>,
Vladimir Medvedkin <vladimir.medvedkin@intel.com>
Subject: Re: [PATCH v3] hash: put GFNI stubs back
Date: Thu, 7 Mar 2024 12:05:33 +0100 [thread overview]
Message-ID: <CAJFAV8wsYRZFan_Qo_wqdRaRFT8wY-3m80TWgiwd2C=uyL_6tA@mail.gmail.com> (raw)
In-Reply-To: <20240307013715.114013-1-stephen@networkplumber.org>
Hello,
On Thu, Mar 7, 2024 at 2:37 AM Stephen Hemminger
<stephen@networkplumber.org> wrote:
>
> Make the GFNI stub functions always built. This solves the conditional
> linking problem. If GFNI is available, they will never get used.
>
> Fixes: 07d836e5929d ("hash: uninline GFNI stubs")
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
> ---
> v3 - fix typo in call to ___rte_thash_gfni_bulk
>
> lib/hash/rte_thash_gfni.c | 9 +++------
> lib/hash/rte_thash_gfni.h | 30 +++++++++++++++++++++++-------
> lib/hash/version.map | 9 +++++++--
> 3 files changed, 33 insertions(+), 15 deletions(-)
>
> diff --git a/lib/hash/rte_thash_gfni.c b/lib/hash/rte_thash_gfni.c
> index f1525f9838de..5ead51dd3426 100644
> --- a/lib/hash/rte_thash_gfni.c
> +++ b/lib/hash/rte_thash_gfni.c
> @@ -4,18 +4,17 @@
>
> #include <stdbool.h>
>
> +#include <rte_compat.h>
This should be in the header since this is where the __rte_internal tag is used.
Moving this inclusion will fix the failing headers check for arches != x86:
http://mails.dpdk.org/archives/test-report/2024-March/603562.html
> #include <rte_log.h>
> #include <rte_thash_gfni.h>
>
> -#ifndef RTE_THASH_GFNI_DEFINED
> -
> RTE_LOG_REGISTER_SUFFIX(hash_gfni_logtype, gfni, INFO);
> #define RTE_LOGTYPE_HASH hash_gfni_logtype
> #define HASH_LOG(level, ...) \
> RTE_LOG_LINE(level, HASH, "" __VA_ARGS__)
>
> uint32_t
> -rte_thash_gfni(const uint64_t *mtrx __rte_unused,
> +___rte_thash_gfni(const uint64_t *mtrx __rte_unused,
Let's go with a s/___rte_thash_gfni/rte_thash_gfni_stub/g.
> const uint8_t *key __rte_unused, int len __rte_unused)
> {
> static bool warned;
> @@ -30,7 +29,7 @@ rte_thash_gfni(const uint64_t *mtrx __rte_unused,
> }
>
> void
> -rte_thash_gfni_bulk(const uint64_t *mtrx __rte_unused,
> +___rte_thash_gfni_bulk(const uint64_t *mtrx __rte_unused,
> int len __rte_unused, uint8_t *tuple[] __rte_unused,
> uint32_t val[], uint32_t num)
> {
> @@ -47,5 +46,3 @@ rte_thash_gfni_bulk(const uint64_t *mtrx __rte_unused,
> for (i = 0; i < num; i++)
> val[i] = 0;
> }
> -
> -#endif
> diff --git a/lib/hash/rte_thash_gfni.h b/lib/hash/rte_thash_gfni.h
> index eed55fc86c86..432d5b63a12e 100644
> --- a/lib/hash/rte_thash_gfni.h
> +++ b/lib/hash/rte_thash_gfni.h
> @@ -17,11 +17,22 @@ extern "C" {
>
> #endif
>
> -#ifndef RTE_THASH_GFNI_DEFINED
> +/*
> + * @internal
> + * Stubs defined for only used when GFNI is not available
This reads strange, I'll reword when applying as:
* Stubs only used when GFNI is not available
> + */
> +__rte_internal
> +uint32_t
> +___rte_thash_gfni(const uint64_t *mtrx, const uint8_t *key, int len);
> +
> +__rte_internal
> +void
> +___rte_thash_gfni_bulk(const uint64_t *mtrx, int len, uint8_t *tuple[],
> + uint32_t val[], uint32_t num);
>
> +#ifndef RTE_THASH_GFNI_DEFINED
> /**
> * Calculate Toeplitz hash.
> - * Dummy implementation.
> *
> * @param m
> * Pointer to the matrices generated from the corresponding
> @@ -33,12 +44,14 @@ extern "C" {
> * @return
> * Calculated Toeplitz hash value.
> */
> -uint32_t
> -rte_thash_gfni(const uint64_t *mtrx, const uint8_t *key, int len);
> +static inline uint32_t
> +rte_thash_gfni(const uint64_t *mtrx, const uint8_t *key, int len)
> +{
> + return ___rte_thash_gfni(mtrx, key, len);
> +}
>
> /**
> * Bulk implementation for Toeplitz hash.
> - * Dummy implementation.
> *
> * @param m
> * Pointer to the matrices generated from the corresponding
> @@ -53,9 +66,12 @@ rte_thash_gfni(const uint64_t *mtrx, const uint8_t *key, int len);
> * @param num
> * Number of tuples to hash.
> */
> -void
> +static inline void
> rte_thash_gfni_bulk(const uint64_t *mtrx, int len, uint8_t *tuple[],
> - uint32_t val[], uint32_t num);
> + uint32_t val[], uint32_t num)
> +{
> + return ___rte_thash_gfni_bulk(mtrx, len, tuple, val, num);
> +}
>
> #endif /* RTE_THASH_GFNI_DEFINED */
>
> diff --git a/lib/hash/version.map b/lib/hash/version.map
> index 6b2afebf6b46..942e2998578f 100644
> --- a/lib/hash/version.map
> +++ b/lib/hash/version.map
> @@ -41,10 +41,15 @@ DPDK_24 {
> rte_thash_get_gfni_matrices;
> rte_thash_get_helper;
> rte_thash_get_key;
> - rte_thash_gfni;
> - rte_thash_gfni_bulk;
> rte_thash_gfni_supported;
> rte_thash_init_ctx;
>
> local: *;
> };
> +
> +INTERNAL {
> + global:
> +
> + ___rte_thash_gfni;
> + ___rte_thash_gfni_bulk;
> +};
> --
> 2.43.0
>
The rest lgtm but I can't reproduce a link error in the first place.
I'll wait for Tyler to confirm this patch fixes the issue he hit.
--
David Marchand
next prev parent reply other threads:[~2024-03-07 11:05 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-04 18:45 [PATCH] hash: make gfni stubs inline Stephen Hemminger
2024-03-05 3:07 ` [PATCH v2] hash: make GFNI stubs inline (again) Stephen Hemminger
2024-03-05 3:58 ` Tyler Retzlaff
2024-03-06 17:22 ` Thomas Monjalon
2024-03-05 10:14 ` [PATCH] hash: make gfni stubs inline David Marchand
2024-03-05 17:53 ` Tyler Retzlaff
2024-03-05 18:44 ` Stephen Hemminger
2024-03-07 10:32 ` David Marchand
2024-03-07 16:49 ` Stephen Hemminger
2024-03-06 21:47 ` [PATCH v3] hash: put GFNI stubs back Stephen Hemminger
2024-03-07 1:36 ` Stephen Hemminger
2024-03-07 11:05 ` David Marchand [this message]
2024-03-07 17:36 ` Tyler Retzlaff
2024-03-07 17:59 ` David Marchand
2024-03-07 20:23 ` Stephen Hemminger
2024-03-07 19:14 ` [PATCH v4] " Stephen Hemminger
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAJFAV8wsYRZFan_Qo_wqdRaRFT8wY-3m80TWgiwd2C=uyL_6tA@mail.gmail.com' \
--to=david.marchand@redhat.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=roretzla@linux.microsoft.com \
--cc=sameh.gobriel@intel.com \
--cc=stephen@networkplumber.org \
--cc=vladimir.medvedkin@intel.com \
--cc=yipeng1.wang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).