From: "Juraj Linkeš" <juraj.linkes@pantheon.tech>
To: Pavan Nikhilesh Bhagavatula <pbhagavatula@marvell.com>
Cc: Jerin Jacob <jerinj@marvell.com>,
"Ruifeng.Wang@arm.com" <Ruifeng.Wang@arm.com>,
"nd@arm.com" <nd@arm.com>,
Bruce Richardson <bruce.richardson@intel.com>,
"dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [EXT] Re: [PATCH 1/2] config/arm: avoid mcpu and march conflicts
Date: Wed, 31 Jan 2024 10:03:40 +0100 [thread overview]
Message-ID: <CAOb5WZaihO8n1L6tgz-3WFUifZwA-AaYBK5L_oN7jZV8RNsUwA@mail.gmail.com> (raw)
In-Reply-To: <PH0PR18MB40861741D9BE1F6D276C1253DE7D2@PH0PR18MB4086.namprd18.prod.outlook.com>
On Tue, Jan 30, 2024 at 5:16 PM Pavan Nikhilesh Bhagavatula
<pbhagavatula@marvell.com> wrote:
>
>
>
> > -----Original Message-----
> > From: Juraj Linkeš <juraj.linkes@pantheon.tech>
> > Sent: Monday, January 29, 2024 2:15 PM
> > To: Pavan Nikhilesh Bhagavatula <pbhagavatula@marvell.com>
> > Cc: Jerin Jacob <jerinj@marvell.com>; Ruifeng.Wang@arm.com;
> > nd@arm.com; Bruce Richardson <bruce.richardson@intel.com>;
> > dev@dpdk.org
> > Subject: Re: [EXT] Re: [PATCH 1/2] config/arm: avoid mcpu and march
> > conflicts
> >
> > On Wed, Jan 24, 2024 at 4:22 PM Pavan Nikhilesh Bhagavatula
> > <pbhagavatula@marvell.com> wrote:
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: Juraj Linkeš <juraj.linkes@pantheon.tech>
> > > > Sent: Monday, January 22, 2024 9:57 PM
> > > > To: Pavan Nikhilesh Bhagavatula <pbhagavatula@marvell.com>
> > > > Cc: Jerin Jacob <jerinj@marvell.com>; Ruifeng.Wang@arm.com;
> > > > nd@arm.com; Bruce Richardson <bruce.richardson@intel.com>;
> > > > dev@dpdk.org
> > > > Subject: Re: [EXT] Re: [PATCH 1/2] config/arm: avoid mcpu and march
> > > > conflicts
> > > >
> > > > On Mon, Jan 22, 2024 at 12:54 PM Pavan Nikhilesh Bhagavatula
> > > > <pbhagavatula@marvell.com> wrote:
> > > > >
> > > > > > On Sun, Jan 21, 2024 at 10:37 AM <pbhagavatula@marvell.com>
> > wrote:
> > > > > > >
> > > > > > > From: Pavan Nikhilesh <pbhagavatula@marvell.com>
> > > > > > >
> > > > > > > The compiler options march and mtune are a subset
> > > > > > > of mcpu and will lead to conflicts if improper march
> > > > > > > is chosen for a given mcpu.
> > > > > > > To avoid conflicts, force part number march when
> > > > > > > mcpu is available and is supported by the compiler.
> > > > > > >
> > > > > > > Example:
> > > > > > > march = armv9-a
> > > > > > > mcpu = neoverse-n2
> > > > > > >
> > > > > > > mcpu supported, march supported
> > > > > > > machine_args = ['-mcpu=neoverse-n2', '-march=armv9-a']
> > > > > > >
> > > > > > > mcpu supported, march not supported
> > > > > > > machine_args = ['-mcpu=neoverse-n2']
> > > > > > >
> > > > > > > mcpu not supported, march supported
> > > > > > > machine_args = ['-march=armv9-a']
> > > > > > >
> > > > > > > mcpu not supported, march not supported
> > > > > > > machine_args = ['-march=armv8.6-a']
> > > > > > >
> > > > > > > Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
> > > > > > > ---
> > > > > > > config/arm/meson.build | 109 +++++++++++++++++++++++++-----
> > ----
> > > > -----
> > > > > > --
> > > > > > > 1 file changed, 67 insertions(+), 42 deletions(-)
> > > > > > >
> > > > > > > diff --git a/config/arm/meson.build b/config/arm/meson.build
> > > > > > > index 36f21d2259..8c8cfccca0 100644
> > > > > > > --- a/config/arm/meson.build
> > > > > > > +++ b/config/arm/meson.build
> > > > > > <snip>
> > > > > > > @@ -127,21 +128,22 @@ implementer_cavium = {
> > > > > > > ],
> > > > > > > 'part_number_config': {
> > > > > > > '0xa1': {
> > > > > > > - 'compiler_options': ['-mcpu=thunderxt88'],
> > > > > > > + 'mcpu': 'thunderxt88',
> > > > > > > 'flags': flags_part_number_thunderx
> > > > > > > },
> > > > > > > '0xa2': {
> > > > > > > - 'compiler_options': ['-mcpu=thunderxt81'],
> > > > > > > + 'mcpu': 'thunderxt81',
> > > > > > > 'flags': flags_part_number_thunderx
> > > > > > > },
> > > > > > > '0xa3': {
> > > > > > > - 'compiler_options': ['-march=armv8-a+crc', '-
> > > > mcpu=thunderxt83'],
> > > > > > > + 'mcpu': 'thunderxt83',
> > > > > > > + 'compiler_options': ['-march=armv8-a+crc'],
> > > > > >
> > > > > > Let's unify this with the rest and specify 'march': 'armv8-a+crc'
> > > > > > instead of having it under compiler_options.
> > > > >
> > > > > Ack.
> > > > >
> > > > > >
> > > > > > > 'flags': flags_part_number_thunderx
> > > > > > > },
> > > > > > > '0xaf': {
> > > > > > > 'march': 'armv8.1-a',
> > > > > > > 'march_features': ['crc', 'crypto'],
> > > > > > > - 'compiler_options': ['-mcpu=thunderx2t99'],
> > > > > > > + 'mcpu': 'thunderx2t99',
> > > > > > > 'flags': [
> > > > > > > ['RTE_MACHINE', '"thunderx2"'],
> > > > > > > ['RTE_ARM_FEATURE_ATOMICS', true],
> > > > > > > @@ -153,7 +155,7 @@ implementer_cavium = {
> > > > > > > '0xb2': {
> > > > > > > 'march': 'armv8.2-a',
> > > > > > > 'march_features': ['crc', 'crypto', 'lse'],
> > > > > > > - 'compiler_options': ['-mcpu=octeontx2'],
> > > > > > > + 'mcpu': 'octeontx2',
> > > > > > > 'flags': [
> > > > > > > ['RTE_MACHINE', '"cn9k"'],
> > > > > > > ['RTE_ARM_FEATURE_ATOMICS', true],
> > > > > > > @@ -176,7 +178,7 @@ implementer_ampere = {
> > > > > > > '0x0': {
> > > > > > > 'march': 'armv8-a',
> > > > > > > 'march_features': ['crc', 'crypto'],
> > > > > > > - 'compiler_options': ['-mtune=emag'],
> > > > > > > + 'mcpu': 'emag',
> > > > > >
> > > > > > We're changing mtune to mcpu, is this equivalent?
> > > > > >
> > > > >
> > > > > Both march and mtune are a subset of mcpu.
> > > > >
> > > >
> > > > Sure, but we replaced '-mtune=emag' with '-mcpu=emag'. Are these two
> > > > builds going to be different or the same?
> > > >
> > >
> > > Yeah, I believe both are same.
> > >
> >
> > Ok, the change is fine then.
> >
> > > > > > > 'flags': [
> > > > > > > ['RTE_MACHINE', '"eMAG"'],
> > > > > > > ['RTE_MAX_LCORE', 32],
> > > > > > > @@ -186,7 +188,7 @@ implementer_ampere = {
> > > > > > > '0xac3': {
> > > > > > > 'march': 'armv8.6-a',
> > > > > > > 'march_features': ['crc', 'crypto'],
> > > > > > > - 'compiler_options': ['-mcpu=ampere1'],
> > > > > > > + 'mcpu': 'ampere1',
> > > > > > > 'flags': [
> > > > > > > ['RTE_MACHINE', '"AmpereOne"'],
> > > > > > > ['RTE_MAX_LCORE', 320],
> > > > > > > @@ -206,7 +208,7 @@ implementer_hisilicon = {
> > > > > > > '0xd01': {
> > > > > > > 'march': 'armv8.2-a',
> > > > > > > 'march_features': ['crypto'],
> > > > > > > - 'compiler_options': ['-mtune=tsv110'],
> > > > > > > + 'mcpu': 'tsv110',
> > > > > > > 'flags': [
> > > > > > > ['RTE_MACHINE', '"Kunpeng 920"'],
> > > > > > > ['RTE_ARM_FEATURE_ATOMICS', true],
> > > > > > > @@ -695,11 +697,23 @@ if update_flags
> > > > > > >
> > > > > > > machine_args = [] # Clear previous machine args
> > > > > > >
> > > > > > > + candidate_mcpu = ''
> > > > > > > + support_mcpu = false
> > > > > > > + if part_number_config.has_key('mcpu')
> > > > > > > + mcpu = part_number_config['mcpu']
> > > > > > > + if (cc.has_argument('-mcpu=' + mcpu))
> > > > > > > + candidate_mcpu = mcpu
> > > > > > > + support_mcpu = true
> > > > > > > + endif
> > > > > > > + endif
> > > > > > > +
> > > > > > > # probe supported archs and their features
> > > > > > > candidate_march = ''
> > > > > > > if part_number_config.has_key('march')
> > > > > > > - if part_number_config.get('force_march', false)
> > > > > > > - candidate_march = part_number_config['march']
> > > > > > > + if part_number_config.get('force_march', false) or
> > support_mcpu
> > > > > >
> > > > > > Instead of using the extra "support_mcpu" variable, we could do the
> > > > > > same check as with candidate march (if candidate_mcpu != '', which we
> > > > > > actually do below in the last lines of the patch).
> > > > > >
> > > > >
> > > > > Ack.
> > > > >
> > > > > > If I understand the logic correctly, we don't want to do the march
> > > > > > fallback if mcpu is specified - either the march works with the given
> > > > > > mcpu or we do without it (because we don't actually need it with
> > > > > > mcpu). Is that correct?
> > > > > >
> > > > >
> > > > > Yes, but still exact march defined in part_number_config should be
> > present
> > > > for setting extra_march_features.
> > > > > specially for expressing crypto support.
> > > > >
> > > >
> > > > Ok, thanks.
> > > >
> > > > > > > + if cc.has_argument('-march=' + part_number_config['march'])
> > > > > >
> > > > > > Now that we've added mcpu into the mix, is this still the right
> > > > > > condition? Can the below happen?
> > > > > >
> > > > > > This check finds that machine_args = ['-march=armv9-a'] is supported.
> > > > > >
> > > > > > But taken together with mcpu (machine_args = ['-mcpu=neoverse-n2',
> > > > > > '-march=armv9-a']), it is not supported? In this case we'll end up
> > > > > > with invalid configuration.
> > > > >
> > > > > This is the only correct option and evolves into -march=armv9-
> > > > a+sve2+crypto for cn10k
> > > > > whereas other neoverse-n2 might only have -march=armv9-a+sve2.
> > > > >
> > > >
> > > > Maybe I should rephrase my question a bit:
> > > >
> > > > The correct options are ['-mcpu=neoverse-n2', '-march=armv9-a']. Is it
> > > > possible that the compiler will say:
> > > >
> > > > ['-mcpu=neoverse-n2', '-march=armv9-a'] is supported
> > > > ['-mcpu=neoverse-n2'] is supported
> > > > ['-march=armv9-a'] is not supported
> > > >
> > >
> > > Yes, this is possible.
> > >
> >
> > Ok, do we want to make sure it's possible to configure the options in
> > this way? But maybe that's out of scope of this patch.
> >
>
> This patch already handles those conditions, this is the case 2 mentioned in the example.
>
> march = armv9-a
> mcpu = neoverse-n2
>
> [1] mcpu supported, march supported
> machine_args = ['-mcpu=neoverse-n2', '-march=armv9-a']
>
> [2] mcpu supported, march not supported
> machine_args = ['-mcpu=neoverse-n2']
>
> [3] mcpu not supported, march supported
> machine_args = ['-march=armv9-a']
>
> [4] mcpu not supported, march not supported
> machine_args = ['-march=armv8.6-a']
>
Ok, that sounds great, thanks.
> > > > So basically the question is are we risking that the compiler will say
> > > > it supports both options only when both are passed while also saying
> > > > it doesn't support one or both of them when checked alone. We've seen
> > > > this behavior with newer compilers in aarch32 builds
> > > > (-march=armv8-a+simd -mfpu=auto are supported when both are passed,
> > > > but -march=armv8-a is not supported alone), so I wanted to be sure.
> > > >
> > > > > Example:
> > > > >
> > > > > Good:
> > > > > #aarch64-linux-gnu-gcc -march=armv9-a+sve2+crypto -mcpu=neoverse-
> > n2
> > > > shrn.c
> > > > > #aarch64-linux-gnu-gcc -march=armv9-a+sve2 -mcpu=neoverse-n2
> > shrn.c
> > > > > #aarch64-linux-gnu-gcc -march=armv9-a -mcpu=neoverse-n2 shrn.c
> > > > >
> > > > > Bad:
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-march=armv8-
> > a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.1-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.1-a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.2-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.2-a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.3-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.3-a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.4-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.4-a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.5-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.5-a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.6-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.6-a'
> > > > switch
> > > > > #aarch64-linux-gnu-gcc -mcpu=neoverse-n2 -march=armv8.7-a shrn.c
> > > > > cc1: warning: switch '-mcpu=neoverse-n2' conflicts with '-
> > march=armv8.7-a'
> > > > switch
> > > > >
> > > > > >
> > > > > > > + candidate_march = part_number_config['march']
> > > > > > > + endif
> > > > > > > else
> > > > > > > supported_marchs = ['armv8.6-a', 'armv8.5-a', 'armv8.4-a',
> > > > 'armv8.3-
> > > > > > a',
> > > > > > > 'armv8.2-a', 'armv8.1-a', 'armv8-a']
next prev parent reply other threads:[~2024-01-31 9:03 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-21 9:36 pbhagavatula
2024-01-21 9:36 ` [PATCH 2/2] config/arm: add support for fallback march pbhagavatula
2024-01-22 6:30 ` Ruifeng Wang
2024-01-22 11:04 ` Juraj Linkeš
2024-01-22 12:16 ` [EXT] " Pavan Nikhilesh Bhagavatula
2024-01-22 16:29 ` Juraj Linkeš
2024-01-24 15:25 ` Pavan Nikhilesh Bhagavatula
2024-01-22 6:29 ` [PATCH 1/2] config/arm: avoid mcpu and march conflicts Ruifeng Wang
2024-01-22 10:55 ` Juraj Linkeš
2024-01-22 11:54 ` [EXT] " Pavan Nikhilesh Bhagavatula
2024-01-22 16:26 ` Juraj Linkeš
2024-01-24 15:21 ` Pavan Nikhilesh Bhagavatula
2024-01-29 8:44 ` Juraj Linkeš
2024-01-30 16:16 ` Pavan Nikhilesh Bhagavatula
2024-01-31 9:03 ` Juraj Linkeš [this message]
2024-02-01 21:57 ` [PATCH v2 1/3] " pbhagavatula
2024-02-01 21:57 ` [PATCH v2 2/3] config/arm: add support for fallback march pbhagavatula
2024-02-01 21:57 ` [PATCH v2 3/3] config/arm: allow WFE to be enabled config time pbhagavatula
2024-02-07 2:55 ` Honnappa Nagarahalli
2024-02-10 6:47 ` Pavan Nikhilesh Bhagavatula
2024-02-10 16:36 ` Honnappa Nagarahalli
2024-02-10 16:40 ` Pavan Nikhilesh Bhagavatula
2024-02-02 8:50 ` [PATCH v3 1/3] config/arm: avoid mcpu and march conflicts pbhagavatula
2024-02-02 8:50 ` [PATCH v3 2/3] config/arm: add support for fallback march pbhagavatula
2024-02-07 20:24 ` Wathsala Wathawana Vithanage
2024-02-02 8:50 ` [PATCH v3 3/3] config/arm: allow WFE to be enabled config time pbhagavatula
2024-02-10 16:56 ` Honnappa Nagarahalli
2024-02-12 19:21 ` Wathsala Wathawana Vithanage
2024-02-06 4:10 ` [PATCH v3 1/3] config/arm: avoid mcpu and march conflicts Wathsala Wathawana Vithanage
2024-02-06 4:44 ` Honnappa Nagarahalli
2024-02-06 10:21 ` Pavan Nikhilesh Bhagavatula
2024-02-07 0:01 ` Wathsala Wathawana Vithanage
2024-02-10 6:49 ` Pavan Nikhilesh Bhagavatula
2024-02-10 15:20 ` Honnappa Nagarahalli
2024-02-10 17:21 ` Honnappa Nagarahalli
2024-02-21 20:20 ` [PATCH v4 " pbhagavatula
2024-02-21 20:20 ` [PATCH v4 2/3] config/arm: add support for fallback march pbhagavatula
2024-02-22 10:47 ` Juraj Linkeš
2024-02-22 12:32 ` [EXT] " Pavan Nikhilesh Bhagavatula
2024-02-21 20:20 ` [PATCH v4 3/3] config/arm: allow WFE to be enabled config time pbhagavatula
2024-02-22 9:37 ` [PATCH v4 1/3] config/arm: avoid mcpu and march conflicts Juraj Linkeš
2024-02-22 9:49 ` [EXT] " Pavan Nikhilesh Bhagavatula
2024-02-22 12:45 ` [PATCH v5 " pbhagavatula
2024-02-22 12:45 ` [PATCH v5 2/3] config/arm: add support for fallback march pbhagavatula
2024-02-23 11:49 ` Juraj Linkeš
2024-02-26 7:11 ` [EXT] " Pavan Nikhilesh Bhagavatula
2024-02-26 7:31 ` Juraj Linkeš
2024-02-26 7:34 ` Juraj Linkeš
2024-02-22 12:45 ` [PATCH v5 3/3] config/arm: allow WFE to be enabled config time pbhagavatula
2024-02-23 11:19 ` [PATCH v5 1/3] config/arm: avoid mcpu and march conflicts Juraj Linkeš
2024-02-26 7:08 ` [EXT] " Pavan Nikhilesh Bhagavatula
2024-02-26 7:38 ` [PATCH v6 " pbhagavatula
2024-02-26 7:38 ` [PATCH v6 2/3] config/arm: add support for fallback march pbhagavatula
2024-02-26 7:38 ` [PATCH v6 3/3] config/arm: allow WFE to be enabled config time pbhagavatula
2024-03-06 15:49 ` [PATCH v7 1/3] config/arm: avoid mcpu and march conflicts pbhagavatula
2024-03-06 15:49 ` [PATCH v7 2/3] config/arm: add support for fallback march pbhagavatula
2024-03-06 15:49 ` [PATCH v7 3/3] config/arm: allow WFE to be enabled config time pbhagavatula
2024-03-07 3:57 ` [PATCH v7 1/3] config/arm: avoid mcpu and march conflicts Pavan Nikhilesh Bhagavatula
2024-03-14 11:38 ` [PATCH v8 1/5] " pbhagavatula
2024-03-14 11:38 ` [PATCH v8 2/5] config/arm: add armv9-a march support pbhagavatula
2024-03-14 11:38 ` [PATCH v8 3/5] config/arm: add crypto march feature to thunderxt83 pbhagavatula
2024-03-14 12:00 ` Jerin Jacob
2024-03-14 11:38 ` [PATCH v8 4/5] config/arm: add support for fallback march pbhagavatula
2024-03-14 11:38 ` [PATCH v8 5/5] config/arm: allow WFE to be enabled config time pbhagavatula
2024-03-15 11:17 ` [PATCH v8 1/5] config/arm: avoid mcpu and march conflicts Thomas Monjalon
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=CAOb5WZaihO8n1L6tgz-3WFUifZwA-AaYBK5L_oN7jZV8RNsUwA@mail.gmail.com \
--to=juraj.linkes@pantheon.tech \
--cc=Ruifeng.Wang@arm.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=jerinj@marvell.com \
--cc=nd@arm.com \
--cc=pbhagavatula@marvell.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).