From: "Daly, Lee" <lee.daly@intel.com>
To: Akhil Goyal <akhil.goyal@nxp.com>,
"Cel, TomaszX" <tomaszx.cel@intel.com>,
"dev@dpdk.org" <dev@dpdk.org>
Cc: "stable@dpdk.org" <stable@dpdk.org>,
"Trahe, Fiona" <fiona.trahe@intel.com>
Subject: Re: [dpdk-stable] [dpdk-dev] [PATCH] compress/isal: fix getting information about CPU
Date: Wed, 27 Mar 2019 15:01:26 +0000 [thread overview]
Message-ID: <F5C6929789601049BEB7272E26735598893455@IRSMSX106.ger.corp.intel.com> (raw)
In-Reply-To: <bd24b0fb-f548-738a-b634-717c09a6f9af@nxp.com>
Hi Tomasz,
This patch makes sense to check does the CPU have the capability of certain instructions before
Adding it to the dev_info flags. I think one more addition should be made in isal_compress_pmd.c
When checking what compression level should be used , decided by the AVX instructions available, it previously used the
Rte_cpu_get_flag_enabled() function, however, I don’t believe that is needed any more since it will now be checked beforehand, it would be doing twice the work.
When the driver now needs to decide the compression level it should now just check the feature_flags set below. Perhaps a V2 could be sent with this addition.
Regards,
Lee.
> > This patch adds query about CPU features
> >
> > Fixes: 53a9baa98c36 ("compress/isal: add basic PMD ops")
> > Cc: stable@dpdk.org
> >
> > Signed-off-by: Tomasz Cel <tomaszx.cel@intel.com>
> > ---
> > drivers/compress/isal/isal_compress_pmd_ops.c | 16 ++++++++++++----
> > 1 file changed, 12 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/compress/isal/isal_compress_pmd_ops.c
> b/drivers/compress/isal/isal_compress_pmd_ops.c
> > index 7b91849..fe99959 100644
> > --- a/drivers/compress/isal/isal_compress_pmd_ops.c
> > +++ b/drivers/compress/isal/isal_compress_pmd_ops.c
> > @@ -135,10 +135,18 @@ isal_comp_pmd_info_get(struct
> rte_compressdev *dev __rte_unused,
> > {
> > if (dev_info != NULL) {
> > dev_info->capabilities = isal_pmd_capabilities;
> > - dev_info->feature_flags = RTE_COMPDEV_FF_CPU_AVX512 |
> > - RTE_COMPDEV_FF_CPU_AVX2 |
> > - RTE_COMPDEV_FF_CPU_AVX |
> > - RTE_COMPDEV_FF_CPU_SSE;
> > +
> > + /* Check CPU for supported vector instruction and set
> > + * feature_flags
> > + */
> > + if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F))
> > + dev_info->feature_flags |=
> RTE_COMPDEV_FF_CPU_AVX512;
> > + else if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2))
> > + dev_info->feature_flags |=
> RTE_COMPDEV_FF_CPU_AVX2;
> > + else if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX))
> > + dev_info->feature_flags |=
> RTE_COMPDEV_FF_CPU_AVX;
> > + else
> > + dev_info->feature_flags |=
> RTE_COMPDEV_FF_CPU_SSE;
> > }
> > }
> >
next prev parent reply other threads:[~2019-03-27 15:02 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-07 11:28 [dpdk-stable] " Tomasz Cel
2019-03-27 14:05 ` [dpdk-stable] [dpdk-dev] " Akhil Goyal
2019-03-27 15:01 ` Daly, Lee [this message]
2019-03-29 15:28 ` Daly, Lee
2019-03-29 15:37 ` Akhil Goyal
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=F5C6929789601049BEB7272E26735598893455@IRSMSX106.ger.corp.intel.com \
--to=lee.daly@intel.com \
--cc=akhil.goyal@nxp.com \
--cc=dev@dpdk.org \
--cc=fiona.trahe@intel.com \
--cc=stable@dpdk.org \
--cc=tomaszx.cel@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).