From: Andy Voltz <andy.voltz@timesys.com>
To: "openembedded-core@lists.openembedded.org"
<openembedded-core@lists.openembedded.org>
Subject: Re: [PATCH] tune-cortexa5.inc: add tune file for cortex-a5, based on a7
Date: Tue, 4 Jun 2013 15:57:26 -0400 [thread overview]
Message-ID: <20130604195725.GH18597@timesys.com> (raw)
In-Reply-To: <mailman.5041.1370320509.3521.openembedded-core@lists.openembedded.org>
> Date: Mon, 03 Jun 2013 23:18:24 +0200
> From: Marcin Juszkiewicz <marcin@juszkiewicz.com.pl>
> To: openembedded-core@lists.openembedded.org
> Subject: Re: [OE-core] [PATCH] tune-cortexa5.inc: add tune file for
> cortex-a5, based on a7
> Message-ID: <51AD0820.8030302@juszkiewicz.com.pl>
> Content-Type: text/plain; charset=UTF-8
>
> W dniu 03.06.2013 18:40, Andy Voltz pisze:
> > Signed-off-by: Andy Voltz <andy.voltz@timesys.com>
> > ---
> > meta/conf/machine/include/tune-cortexa5.inc | 36 +++++++++++++++++++++++++++
>
> I think that it is time to create simple tune-cortex.inc which would be
> used by cores tunes to set all of it.
>
> So tune-cortexa5.inc would be:
>
> ----
> CORTEX_ID = "a5"
> CORTEX_NAME = "A5"
>
> require conf/machine/include/arm/tune-cortex.inc
> ----
>
> and tune-cortex.inc will have things like:
>
> ----
> DEFAULTTUNE ?= "armv7a-neon"
>
> require conf/machine/include/arm/arch-armv7a.inc
>
> TUNEVALID[cortex${CORTEX_ID}] = "Enable Cortex-${CORTEX_NAME} specific
> processor optimizations"
> TUNE_CCARGS .= "${@bb.utils.contains("TUNE_FEATURES",
> "cortex${CORTEX_ID}", " -mtune=cortex-${CORTEX_ID}", "", d)}"
>
> # Little Endian base configs
> AVAILTUNES += "cortex${CORTEX_ID} cortex${CORTEX_ID}t
> cortex${CORTEX_ID}-neon cortex${CORTEX_ID}t-neon"
> ARMPKGARCH_tune-cortex${CORTEX_ID} = "cortex${CORTEX_ID}"
> ARMPKGARCH_tune-cortex${CORTEX_ID}t = "cortex${CORTEX_ID}"
> ARMPKGARCH_tune-cortex${CORTEX_ID}-neon = "cortex${CORTEX_ID}"
> ARMPKGARCH_tune-cortex${CORTEX_ID}t-neon = "cortex${CORTEX_ID}"
> TUNE_FEATURES_tune-cortex${CORTEX_ID} = "${TUNE_FEATURES_tune-armv7a}
> cortex${CORTEX_ID}"
> TUNE_FEATURES_tune-cortex${CORTEX_ID}t = "${TUNE_FEATURES_tune-armv7at}
> cortex${CORTEX_ID}"
> TUNE_FEATURES_tune-cortex${CORTEX_ID}-neon =
> "${TUNE_FEATURES_tune-armv7a-neon} cortex${CORTEX_ID}"
> TUNE_FEATURES_tune-cortex${CORTEX_ID}t-neon =
> "${TUNE_FEATURES_tune-armv7at-neon} cortex${CORTEX_ID}"
> PACKAGE_EXTRA_ARCHS_tune-cortex${CORTEX_ID} =
> "${PACKAGE_EXTRA_ARCHS_tune-armv7a} cortex${CORTEX_ID}-vfp"
> PACKAGE_EXTRA_ARCHS_tune-cortex${CORTEX_ID}t =
> "${PACKAGE_EXTRA_ARCHS_tune-armv7at} cortex${CORTEX_ID}-vfp
> cortex${CORTEX_ID}t2-vfp"
> PACKAGE_EXTRA_ARCHS_tune-cortex${CORTEX_ID}-neon =
> "${PACKAGE_EXTRA_ARCHS_tune-armv7a-neon} cortex${CORTEX_ID}-vfp
> cortex${CORTEX_ID}-vfp-neon"
> PACKAGE_EXTRA_ARCHS_tune-cortex${CORTEX_ID}t-neon =
> "${PACKAGE_EXTRA_ARCHS_tune-armv7at-neon} cortex${CORTEX_ID}-vfp
> cortex${CORTEX_ID}-vfp-neon cortex${CORTEX_ID}t2-vfp
> cortex${CORTEX_ID}t2-vfp-neon"
>
> [..]
> -----
>
> What do you think?
>
>
I just tried this solution with my vybrid cortex-a5 build. The only issue is a
parse error on ${CORTEX_ID} here:
TUNEVALID[cortex${CORTEX_ID}] = "Enable Cortex-${CORTEX_NAME} specific
I amended the changes so that each tune-cortexa5.inc (for example) looks like:
+CORTEX_ID = "a5"
+
+TUNEVALID[cortexa5] = "Enable Cortex-A5 specific processor optimizations"
+
+require conf/machine/include/tune-cortex.inc
I didn't think it was worth having a CORTEX_NAME if TUNEVALID does not parse..
My only other thought is perhaps tune-cortex.inc should be called
tune-cortexa.inc? It seems like the m series is a different animal.
What do you think?
Regards
--
Andy Voltz
Linux Engineer
Timesys Corporation
next parent reply other threads:[~2013-06-04 19:57 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <mailman.5041.1370320509.3521.openembedded-core@lists.openembedded.org>
2013-06-04 19:57 ` Andy Voltz [this message]
2013-06-04 20:03 ` [PATCH] tune-cortexa5.inc: add tune file for cortex-a5, based on a7 Marcin Juszkiewicz
2013-06-04 21:01 ` Mark Hatle
2013-06-03 16:40 Andy Voltz
2013-06-03 16:51 ` Otavio Salvador
2013-06-03 21:18 ` Marcin Juszkiewicz
2013-06-03 21:51 ` Otavio Salvador
2013-06-03 22:09 ` Mark Hatle
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=20130604195725.GH18597@timesys.com \
--to=andy.voltz@timesys.com \
--cc=openembedded-core@lists.openembedded.org \
/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