| Message ID | 20181210102609.28819-1-fanjinke@hygon.cn |
|---|---|
| State | New, archived |
| Headers |
Received: (qmail 81387 invoked by alias); 10 Dec 2018 10:26:44 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: <libc-alpha.sourceware.org> List-Unsubscribe: <mailto:libc-alpha-unsubscribe-##L=##H@sourceware.org> List-Subscribe: <mailto:libc-alpha-subscribe@sourceware.org> List-Archive: <http://sourceware.org/ml/libc-alpha/> List-Post: <mailto:libc-alpha@sourceware.org> List-Help: <mailto:libc-alpha-help@sourceware.org>, <http://sourceware.org/ml/#faqs> Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 80476 invoked by uid 89); 10 Dec 2018 10:26:43 -0000 Authentication-Results: sourceware.org; auth=none X-Spam-SWARE-Status: No, score=-25.0 required=5.0 tests=BAYES_40, GIT_PATCH_0, GIT_PATCH_1, GIT_PATCH_2, GIT_PATCH_3, SPF_PASS autolearn=ham version=3.3.2 spammy=CoLtd, Co., Ltd, Family, Design X-HELO: spam1.hygon.cn From: fanjinke <fanjinke@hygon.cn> To: <libc-alpha@sourceware.org> CC: <carlos@redhat.com>, fanjinke <fanjinke@hygon.cn> Subject: [PATCH] x86: Add Hygon Dhyana support. Date: Mon, 10 Dec 2018 18:26:09 +0800 Message-ID: <20181210102609.28819-1-fanjinke@hygon.cn> MIME-Version: 1.0 Content-Type: text/plain X-MAIL: spam1.hygon.cn wBAAQMiV079159 X-DNSRBL: |
Commit Message
fanjinke
Dec. 10, 2018, 10:26 a.m. UTC
This patch fix Hygon Dhyana processor CPU Vendor ID detection problem
in glibc sysdep module, current glibc codes doesn't recognize Dhyana
CPU Vendor ID("HygonGenuine") and set kind to arch_kind_other, which
result to incorrect zero value for __cache_sysconf() syscall. As Hygon
Dhyana share most architecture feature as AMD Family 17h, this patch
add Hygon CPU Vendor ID check and setup kind to arch_kind_amd and reuse
AMD code path, which lead to correct return value in __cache_sysconf()
syscall. we run the glibc test suite for both Hygon Dhyana and AMD EPYC
and found no failure case.
Background:
Chengdu Haiguang IC Design Co., Ltd (Hygon) is a Joint Venture between
AMD and Haiguang Information Technology Co.,Ltd., aims at providing high
performance x86 processor for China server market. Its first generation
processor codename is Dhyana, which originates from AMD technology and
shares most of the architecture with AMD's family 17h, but with different
CPU Vendor ID("HygonGenuine")/Family series number(Family 18h).
Related Hygon kernel patch can be found on
http://lkml.kernel.org/r/5ce86123a7b9dad925ac583d88d2f921040e859b.
1538583282.git.puwen@hygon.cn
Very appreciate everyone who has helped review and fix issues.
*Tested for Hygon:*
Before patching the results of hygon's tests:
Summary of test results:
6 FAIL
5927 PASS
29 UNSUPPORTED
17 XFAIL
2 XPASS
Makefile:403: recipe for target 'tests' failed
make[1]: *** [tests] Error 1
make[1]: Leaving directory
'/home/higon/space/src/opensrc/glibc/glibc/glibc'
Makefile:9: recipe for target 'check' failed
make: *** [check] Error 2
And the result of testing after patching:
Summary of test results:
5945 PASS
17 UNSUPPORTED
17 XFAIL
2 XPASS
make[1]: Leaving directory
'/home/higon/space/src/opensrc/glibc/glibc/glibc'
*Tested for AMD:*
The test results did not change before and after patching:
Summary of test results:
5951 PASS
11 UNSUPPORTED
17 XFAIL
2 XPASS
make[1]: Leaving directory '/home/higon/glibc'
ChangeLog:
2018-12-10 fanjinke <fanjinke@hygon.cn>
* sysdeps/x86/cpu-features.c (init_cpu_features):
---
sysdeps/x86/cpu-features.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
Comments
On 12/10/18 5:26 AM, fanjinke wrote: > This patch fix Hygon Dhyana processor CPU Vendor ID detection problem > in glibc sysdep module, current glibc codes doesn't recognize Dhyana > CPU Vendor ID("HygonGenuine") and set kind to arch_kind_other, which > result to incorrect zero value for __cache_sysconf() syscall. As Hygon > Dhyana share most architecture feature as AMD Family 17h, this patch > add Hygon CPU Vendor ID check and setup kind to arch_kind_amd and reuse > AMD code path, which lead to correct return value in __cache_sysconf() > syscall. we run the glibc test suite for both Hygon Dhyana and AMD EPYC > and found no failure case. > > Background: > Chengdu Haiguang IC Design Co., Ltd (Hygon) is a Joint Venture between > AMD and Haiguang Information Technology Co.,Ltd., aims at providing high > performance x86 processor for China server market. Its first generation > processor codename is Dhyana, which originates from AMD technology and > shares most of the architecture with AMD's family 17h, but with different > CPU Vendor ID("HygonGenuine")/Family series number(Family 18h). > > Related Hygon kernel patch can be found on > http://lkml.kernel.org/r/5ce86123a7b9dad925ac583d88d2f921040e859b. > 1538583282.git.puwen@hygon.cn > > Very appreciate everyone who has helped review and fix issues. Do you have copyright assignment papers filed with the FSF? This one-line change doesn't need copyright papers filed, but if you need to make more compilcated changes, then you will need them. > *Tested for Hygon:* > Before patching the results of hygon's tests: > Summary of test results: > 6 FAIL > 5927 PASS > 29 UNSUPPORTED > 17 XFAIL > 2 XPASS > Makefile:403: recipe for target 'tests' failed > make[1]: *** [tests] Error 1 > make[1]: Leaving directory > '/home/higon/space/src/opensrc/glibc/glibc/glibc' > Makefile:9: recipe for target 'check' failed > make: *** [check] Error 2 > > And the result of testing after patching: > Summary of test results: > 5945 PASS > 17 UNSUPPORTED > 17 XFAIL > 2 XPASS > make[1]: Leaving directory > '/home/higon/space/src/opensrc/glibc/glibc/glibc' > > *Tested for AMD:* > The test results did not change before and after patching: > Summary of test results: > 5951 PASS > 11 UNSUPPORTED > 17 XFAIL > 2 XPASS > make[1]: Leaving directory '/home/higon/glibc' > > ChangeLog: > > 2018-12-10 fanjinke <fanjinke@hygon.cn> > > * sysdeps/x86/cpu-features.c (init_cpu_features): > --- > sysdeps/x86/cpu-features.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/sysdeps/x86/cpu-features.c b/sysdeps/x86/cpu-features.c > index 5f9eefd408..86e78bfb78 100644 > --- a/sysdeps/x86/cpu-features.c > +++ b/sysdeps/x86/cpu-features.c > @@ -425,8 +425,9 @@ init_cpu_features (struct cpu_features *cpu_features) > cpu_features->feature[index_arch_Prefer_No_AVX512] > |= bit_arch_Prefer_No_AVX512; > } > - /* This spells out "AuthenticAMD". */ > - else if (ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) > + /* This spells out "AuthenticAMD" or "HygonGenuine". */ > + else if ((ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) > + || (ebx == 0x6f677948 && ecx == 0x656e6975 && edx == 0x6e65476e)) > { > unsigned int extended_model; This looks fine to me, and reuses the AMD code. I'd like to hear from Amit if this is OK for him, since he'll have to maintain this code as things diverge. Amit, Any objections? Reviewed-by: Carlos O'Donell <carlos@redhat.com>
On 12/11/2018 11:15 AM, Carlos O'Donell wrote: > On 12/10/18 5:26 AM, fanjinke wrote: >> This patch fix Hygon Dhyana processor CPU Vendor ID detection problem >> in glibc sysdep module, current glibc codes doesn't recognize Dhyana >> CPU Vendor ID("HygonGenuine") and set kind to arch_kind_other, which >> result to incorrect zero value for __cache_sysconf() syscall. As Hygon >> Dhyana share most architecture feature as AMD Family 17h, this patch >> add Hygon CPU Vendor ID check and setup kind to arch_kind_amd and reuse >> AMD code path, which lead to correct return value in __cache_sysconf() >> syscall. we run the glibc test suite for both Hygon Dhyana and AMD EPYC >> and found no failure case. >> >> Background: >> Chengdu Haiguang IC Design Co., Ltd (Hygon) is a Joint Venture between >> AMD and Haiguang Information Technology Co.,Ltd., aims at providing high >> performance x86 processor for China server market. Its first generation >> processor codename is Dhyana, which originates from AMD technology and >> shares most of the architecture with AMD's family 17h, but with different >> CPU Vendor ID("HygonGenuine")/Family series number(Family 18h). >> >> Related Hygon kernel patch can be found on >> http://lkml.kernel.org/r/5ce86123a7b9dad925ac583d88d2f921040e859b. >> 1538583282.git.puwen@hygon.cn >> >> Very appreciate everyone who has helped review and fix issues. > Do you have copyright assignment papers filed with the FSF? > > This one-line change doesn't need copyright papers filed, but if > you need to make more compilcated changes, then you will need > them. > >> *Tested for Hygon:* >> Before patching the results of hygon's tests: >> Summary of test results: >> 6 FAIL >> 5927 PASS >> 29 UNSUPPORTED >> 17 XFAIL >> 2 XPASS >> Makefile:403: recipe for target 'tests' failed >> make[1]: *** [tests] Error 1 >> make[1]: Leaving directory >> '/home/higon/space/src/opensrc/glibc/glibc/glibc' >> Makefile:9: recipe for target 'check' failed >> make: *** [check] Error 2 >> >> And the result of testing after patching: >> Summary of test results: >> 5945 PASS >> 17 UNSUPPORTED >> 17 XFAIL >> 2 XPASS >> make[1]: Leaving directory >> '/home/higon/space/src/opensrc/glibc/glibc/glibc' >> >> *Tested for AMD:* >> The test results did not change before and after patching: >> Summary of test results: >> 5951 PASS >> 11 UNSUPPORTED >> 17 XFAIL >> 2 XPASS >> make[1]: Leaving directory '/home/higon/glibc' >> >> ChangeLog: >> >> 2018-12-10 fanjinke<fanjinke@hygon.cn> >> >> * sysdeps/x86/cpu-features.c (init_cpu_features): >> --- >> sysdeps/x86/cpu-features.c | 5 +++-- >> 1 file changed, 3 insertions(+), 2 deletions(-) >> >> diff --git a/sysdeps/x86/cpu-features.c b/sysdeps/x86/cpu-features.c >> index 5f9eefd408..86e78bfb78 100644 >> --- a/sysdeps/x86/cpu-features.c >> +++ b/sysdeps/x86/cpu-features.c >> @@ -425,8 +425,9 @@ init_cpu_features (struct cpu_features *cpu_features) >> cpu_features->feature[index_arch_Prefer_No_AVX512] >> |= bit_arch_Prefer_No_AVX512; >> } >> - /* This spells out "AuthenticAMD". */ >> - else if (ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) >> + /* This spells out "AuthenticAMD" or "HygonGenuine". */ >> + else if ((ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) >> + || (ebx == 0x6f677948 && ecx == 0x656e6975 && edx == 0x6e65476e)) >> { >> unsigned int extended_model; > > This looks fine to me, and reuses the AMD code. > > I'd like to hear from Amit if this is OK for him, since he'll have to > maintain this code as things diverge. > > Amit, Any objections? > > Reviewed-by: Carlos O'Donell<carlos@redhat.com> > Hi Carlos, Thank you for your reply and reminder. The copyright issue is already under processing.
> -----Original Message----- > From: Carlos O'Donell <carlos@redhat.com> > Sent: Tuesday, December 11, 2018 8:45 AM > To: fanjinke <fanjinke@hygon.cn>; libc-alpha@sourceware.org; Pawar, Amit > <Amit.Pawar@amd.com> > Cc: H.J. Lu <hjl.tools@gmail.com> > Subject: Re: [PATCH] x86: Add Hygon Dhyana support. > > On 12/10/18 5:26 AM, fanjinke wrote: > > This patch fix Hygon Dhyana processor CPU Vendor ID detection problem > > in glibc sysdep module, current glibc codes doesn't recognize Dhyana > > CPU Vendor ID("HygonGenuine") and set kind to arch_kind_other, which > > result to incorrect zero value for __cache_sysconf() syscall. As Hygon > > Dhyana share most architecture feature as AMD Family 17h, this patch > > add Hygon CPU Vendor ID check and setup kind to arch_kind_amd and > > reuse AMD code path, which lead to correct return value in > > __cache_sysconf() syscall. we run the glibc test suite for both Hygon > > Dhyana and AMD EPYC and found no failure case. > > > > Background: > > Chengdu Haiguang IC Design Co., Ltd (Hygon) is a Joint Venture between > > AMD and Haiguang Information Technology Co.,Ltd., aims at providing > > high performance x86 processor for China server market. Its first > > generation processor codename is Dhyana, which originates from AMD > > technology and shares most of the architecture with AMD's family 17h, > > but with different CPU Vendor ID("HygonGenuine")/Family series > number(Family 18h). > > > > Related Hygon kernel patch can be found on > > http://lkml.kernel.org/r/5ce86123a7b9dad925ac583d88d2f921040e859b. > > 1538583282.git.puwen@hygon.cn > > > > Very appreciate everyone who has helped review and fix issues. > > Do you have copyright assignment papers filed with the FSF? > > This one-line change doesn't need copyright papers filed, but if you need to > make more compilcated changes, then you will need them. > > > *Tested for Hygon:* > > Before patching the results of hygon's tests: > > Summary of test results: > > 6 FAIL > > 5927 PASS > > 29 UNSUPPORTED > > 17 XFAIL > > 2 XPASS > > Makefile:403: recipe for target 'tests' failed > > make[1]: *** [tests] Error 1 > > make[1]: Leaving directory > > '/home/higon/space/src/opensrc/glibc/glibc/glibc' > > Makefile:9: recipe for target 'check' failed > > make: *** [check] Error 2 > > > > And the result of testing after patching: > > Summary of test results: > > 5945 PASS > > 17 UNSUPPORTED > > 17 XFAIL > > 2 XPASS > > make[1]: Leaving directory > > '/home/higon/space/src/opensrc/glibc/glibc/glibc' > > > > *Tested for AMD:* > > The test results did not change before and after patching: > > Summary of test results: > > 5951 PASS > > 11 UNSUPPORTED > > 17 XFAIL > > 2 XPASS > > make[1]: Leaving directory '/home/higon/glibc' > > > > ChangeLog: > > > > 2018-12-10 fanjinke <fanjinke@hygon.cn> > > > > * sysdeps/x86/cpu-features.c (init_cpu_features): > > --- > > sysdeps/x86/cpu-features.c | 5 +++-- > > 1 file changed, 3 insertions(+), 2 deletions(-) > > > > diff --git a/sysdeps/x86/cpu-features.c b/sysdeps/x86/cpu-features.c > > index 5f9eefd408..86e78bfb78 100644 > > --- a/sysdeps/x86/cpu-features.c > > +++ b/sysdeps/x86/cpu-features.c > > @@ -425,8 +425,9 @@ init_cpu_features (struct cpu_features > *cpu_features) > > cpu_features->feature[index_arch_Prefer_No_AVX512] > > |= bit_arch_Prefer_No_AVX512; > > } > > - /* This spells out "AuthenticAMD". */ > > - else if (ebx == 0x68747541 && ecx == 0x444d4163 && edx == > > 0x69746e65) > > + /* This spells out "AuthenticAMD" or "HygonGenuine". */ > > + else if ((ebx == 0x68747541 && ecx == 0x444d4163 && edx == > 0x69746e65) > > + || (ebx == 0x6f677948 && ecx == 0x656e6975 && edx == > 0x6e65476e)) > > { > > unsigned int extended_model; > > This looks fine to me, and reuses the AMD code. > > I'd like to hear from Amit if this is OK for him, since he'll have to maintain this > code as things diverge. > > Amit, Any objections? No objection from my side. > > Reviewed-by: Carlos O'Donell <carlos@redhat.com> > > -- > Cheers, > Carlos. Thanks, Amit Pawar
diff --git a/sysdeps/x86/cpu-features.c b/sysdeps/x86/cpu-features.c index 5f9eefd408..86e78bfb78 100644 --- a/sysdeps/x86/cpu-features.c +++ b/sysdeps/x86/cpu-features.c @@ -425,8 +425,9 @@ init_cpu_features (struct cpu_features *cpu_features) cpu_features->feature[index_arch_Prefer_No_AVX512] |= bit_arch_Prefer_No_AVX512; } - /* This spells out "AuthenticAMD". */ - else if (ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) + /* This spells out "AuthenticAMD" or "HygonGenuine". */ + else if ((ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) + || (ebx == 0x6f677948 && ecx == 0x656e6975 && edx == 0x6e65476e)) { unsigned int extended_model;