Message ID | 20230711201042.38285-1-raj.khem@gmail.com |
---|---|
State | New |
Headers | show |
Series | [1/2] tunes: Add support for sve instructions on armv8/armv9 | expand |
On 11 Jul 2023, at 21:10, Khem Raj via lists.openembedded.org <raj.khem=gmail.com@lists.openembedded.org> wrote: > > This helps in defining correct compiler options and configure options > for glibc and overall ABI for toolchain Given this, shouldn’t the glibc patch respect the tune flags instead of forcibly turning sve off? Ross
On Wed, Jul 12, 2023 at 3:29 AM Ross Burton <Ross.Burton@arm.com> wrote: > > On 11 Jul 2023, at 21:10, Khem Raj via lists.openembedded.org <raj.khem=gmail.com@lists.openembedded.org> wrote: > > > > This helps in defining correct compiler options and configure options > > for glibc and overall ABI for toolchain > > Given this, shouldn’t the glibc patch respect the tune flags instead of forcibly turning sve off? It will as a follow up. > > Ross
Hello Khem, On 12/07/2023 09:12:54-0700, Khem Raj wrote: > On Wed, Jul 12, 2023 at 3:29 AM Ross Burton <Ross.Burton@arm.com> wrote: > > > > On 11 Jul 2023, at 21:10, Khem Raj via lists.openembedded.org <raj.khem=gmail.com@lists.openembedded.org> wrote: > > > > > > This helps in defining correct compiler options and configure options > > > for glibc and overall ABI for toolchain > > > > Given this, shouldn’t the glibc patch respect the tune flags instead of forcibly turning sve off? > > It will as a follow up. I think we are still waiting for the follow up.
On Fri, Jul 21, 2023 at 7:34 AM Alexandre Belloni < alexandre.belloni@bootlin.com> wrote: > Hello Khem, > > On 12/07/2023 09:12:54-0700, Khem Raj wrote: > > On Wed, Jul 12, 2023 at 3:29 AM Ross Burton <Ross.Burton@arm.com> wrote: > > > > > > On 11 Jul 2023, at 21:10, Khem Raj via lists.openembedded.org > <raj.khem=gmail.com@lists.openembedded.org> wrote: > > > > > > > > This helps in defining correct compiler options and configure options > > > > for glibc and overall ABI for toolchain > > > > > > Given this, shouldn’t the glibc patch respect the tune flags instead > of forcibly turning sve off? > > > > It will as a follow up. > > I think we are still waiting for the follow up. Hmm I was thinking of sending it separate since it does not impact this pull I am away this week perhaps will do upon return next week > > > -- > Alexandre Belloni, co-owner and COO, Bootlin > Embedded Linux and Kernel engineering > https://bootlin.com >
diff --git a/meta/conf/machine/include/arm/feature-arm-sve.inc b/meta/conf/machine/include/arm/feature-arm-sve.inc new file mode 100644 index 00000000000..279bf8c4862 --- /dev/null +++ b/meta/conf/machine/include/arm/feature-arm-sve.inc @@ -0,0 +1,5 @@ +# Scalable Vector Extension (SVE) for: +# armv8.2-a, armv8.3-a, armv8.4-a, armv8.5-a, armv8.6-a, armv8-r, armv9a + +TUNEVALID[sve] = "Enable SVE instructions for ARMv8 and ARMv9" +TUNE_CCARGS_MARCH_OPTS .= "${@bb.utils.contains('TUNE_FEATURES', 'sve', '+sve', '', d)}"
This helps in defining correct compiler options and configure options for glibc and overall ABI for toolchain Signed-off-by: Khem Raj <raj.khem@gmail.com> --- meta/conf/machine/include/arm/feature-arm-sve.inc | 5 +++++ 1 file changed, 5 insertions(+) create mode 100644 meta/conf/machine/include/arm/feature-arm-sve.inc