package/mesa3d: fix nouveau std::isinf related compile failure
PATCH v2 1/1 package/mesa3d: fix nouveau std::isinf related compile failure
Activate already existing mesa3d solution for the isinf compile failure for uclibc based toolchains instead of using a custom workaround.
- remove 0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch
- add 0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch
Fixes:
http://autobuild.buildroot.net/results/cbefc5d4a4fefb674e596400fa1d2698cd89c5b3/
http://autobuild.buildroot.net/results/dc974da012f53fa4ed3be616f937b0afae423d66/
../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: In member function 'bool nv50_ir::GCRA::simplify()':
../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1348:19: error: expected unqualified-id before '(' token
if (std::isinf(bestScore)) {
^---
Notes:
- as Romain Naour pointed out ([1], [2]) the existing check in c99_math.h failes for uclibc (because it impersonate glibc-2.2), so enable the original isinf solution for uclibc instead of the custom work around
Changes v1 -> v2:
- update patch to use defined() around __UCLIBC__
- update uclibc impersonate version to 2.2 in notes and fix wording
[1] http://lists.busybox.net/pipermail/buildroot/2020-March/275624.html
[2] http://lists.busybox.net/pipermail/buildroot/2020-March/275735.html
--- ...rt-isinf-for-uclibc-based-toolchains.patch | 30 ++++++++++++ ...ers-nouveau-codegen-nv50_ir_ra.cpp-p.patch | 47 ------------------- 2 files changed, 30 insertions(+), 47 deletions(-) create mode 100644 package/mesa3d/0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch delete mode 100644 package/mesa3d/0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch
package/mesa3d/0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch new file +30 -0
@@ -0,0 +1,30 @@ +From 705490763fcc052defa2d49a5c5200775cd12c54 Mon Sep 17 00:00:00 2001 +From: Peter Seiderer <[email protected]> +Date: Wed, 4 Mar 2020 08:27:24 +0100 +Subject: [PATCH] c99_math: import isinf for uclibc based toolchains + +The same fix as for glibc >= 2.23 is needed for uclibc +based toolchains. + +[Upstream suggested: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4042] +Signed-off-by: Peter Seiderer <[email protected]> +--- + include/c99_math.h | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/include/c99_math.h b/include/c99_math.h +index e906c26aa54..c8a73a2d2c7 100644 +--- a/include/c99_math.h ++++ b/include/c99_math.h +@@ -190,7 +190,7 @@ fpclassify(double x) + * undefines those functions, which in glibc 2.23, are defined as macros rather + * than functions as in glibc 2.22. + */ +-#if __cplusplus >= 201103L && (__GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 23)) ++#if __cplusplus >= 201103L && (__GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 23) || defined(__UCLIBC__)) + #include <cmath> + + using std::fpclassify; +-- +2.25.1 +
package/mesa3d/0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch deleted +0 -47
47 lines not on this page · raw message
Signature · 1 line
2.25.1
PATCH v2 1/1 package/mesa3d: fix nouveau std::isinf related compile failure
Peter, All,
On 2020-03-04 12:00 +0100, Peter Seiderer spake thusly:
Quoted 10 lines
Activate already existing mesa3d solution for the isinf compile failure for uclibc based toolchains instead of using a custom workaround.
- remove 0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch
- add 0004-c99_math-import-isinf-for-uclibc-based-toolchains.patchFixes:
http://autobuild.buildroot.net/results/cbefc5d4a4fefb674e596400fa1d2698cd89c5b3/
http://autobuild.buildroot.net/results/dc974da012f53fa4ed3be616f937b0afae423d66/
You've tagged this patch for next, but the two failures pccured on master.
Where should this patch be applied, then?
Regards,
Yann E. MORIN.
Quoted 123 lines
../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: In member function 'bool
nv50_ir::GCRA::simplify()':
../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1348:19: error: expected unqualified-id before '(' tokenif (std::isinf(bestScore)) { ^Signed-off-by Peter Seiderer <[email protected]>---
Notes:
- as Romain Naour pointed out ([1], [2]) the existing check inc99_math.hfailes for uclibc (because it impersonate glibc-2.2), so enable the original isinf solution for uclibc instead of the custom work aroundChanges v1 -> v2:
- update patch to usedefined()around__UCLIBC__
- update uclibc impersonate version to 2.2 in notes and fix wording[1] http://lists.busybox.net/pipermail/buildroot/2020-March/275624.html
[2] http://lists.busybox.net/pipermail/buildroot/2020-March/275735.html--- ...rt-isinf-for-uclibc-based-toolchains.patch | 30 ++++++++++++ ...ers-nouveau-codegen-nv50_ir_ra.cpp-p.patch | 47 ------------------- 2 files changed, 30 insertions(+), 47 deletions(-) create mode 100644 package/mesa3d/0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch delete mode 100644 package/mesa3d/0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patchpackage/mesa3d/0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch new file +30
-0@@ -0,0 +1,30 @@ +From 705490763fcc052defa2d49a5c5200775cd12c54 Mon Sep 17 00:00:00 2001 +From: Peter Seiderer <[email protected]> +Date: Wed, 4 Mar 2020 08:27:24 +0100 +Subject: [PATCH] c99_math: import isinf for uclibc based toolchains + +The same fix as for glibc >= 2.23 is needed for uclibc +based toolchains. + +[Upstream suggested: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4042] +Signed-off-by: Peter Seiderer <[email protected]> +--- + include/c99_math.h | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/include/c99_math.h b/include/c99_math.h +index e906c26aa54..c8a73a2d2c7 100644 +--- a/include/c99_math.h ++++ b/include/c99_math.h +@@ -190,7 +190,7 @@ fpclassify(double x) + * undefines those functions, which in glibc 2.23, are defined as macros rather + * than functions as in glibc 2.22. + */ +-#if __cplusplus >= 201103L && (__GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 23)) ++#if __cplusplus >= 201103L && (__GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 23) || defined(__UCLIBC__)) + #include <cmath> + + using std::fpclassify; +-- +2.25.1 +package/mesa3d/0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch deleted +0
-47@@ -1,47 +0,0 @@ -From f2ce27b761eeca301a109856f20ed9a8eb06796d Mon Sep 17 00:00:00 2001 -From: Fabrice Fontaine <[email protected]> -Date: Sat, 28 Dec 2019 21:49:43 +0100 -Subject: [PATCH] src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: prefix - isinf with std:: - -Prefix isinf with std:: to fix the following build failure with gcc 4.9: - -../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: In member function 'bool nv50_ir::GCRA::simplify()': -../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1350:29: error: 'isinf' was not declared in this scope - if (isinf(bestScore)) { - ^ -../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1350:29: note: suggested alternative: -In file included from /data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/random:38:0, - from /data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/bits/stl_algo.h:66, - from /data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/algorithm:62, - from ../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:26: -/data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/cmath:621:5: note: 'std::isinf' - isinf(_Tp __x) - ^ - -Fixes: - - http://autobuild.buildroot.org/results/599dc3478ed65d36fbf9f5e9625691dfa813b530 - -Signed-off-by: Fabrice Fontaine <[email protected]> -[Upstream status: -https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3227] ---- - src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp -index 6df2664da22..e81fb117213 100644 ---- a/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp -+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp -@@ -1345,7 +1345,7 @@ GCRA::simplify() - bestMaxReg = it->maxReg; - } - } -- if (isinf(bestScore)) { -+ if (std::isinf(bestScore)) { - ERROR("no viable spill candidates left\n"); - return false; - } --- -2.24.0 -Signature · 6 lines
2.25.1 _______________________________________________ buildroot mailing list buildroot at busybox.net http://lists.busybox.net/mailman/listinfo/buildroot
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------. | Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 561 099 427 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. | '------------------------------^-------^------------------^--------------------'
PATCH v2 1/1 package/mesa3d: fix nouveau std::isinf related compile failure
Hello Yann,
Quoted 10 lines · Yann E. MORIN
Peter, All,
On 2020-03-04 12:00 +0100, Peter Seiderer spake thusly:
Quoted 7 lines
Activate already existing mesa3d solution for the isinf compile failure for uclibc based toolchains instead of using a custom workaround.
- remove 0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch
- add 0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch
Quoted 8 lines
Quoted 3 lines
Fixes:
http://autobuild.buildroot.net/results/cbefc5d4a4fefb674e596400fa1d2698cd89c5b3/
http://autobuild.buildroot.net/results/dc974da012f53fa4ed3be616f937b0afae423d66/You've tagged this patch for next, but the two failures pccured on master.
Where should this patch be applied, then?
The patch [1] was meant for master (minimal change, minimal risk) and this one (v1/v2) for next..., chronological later, first not sure about the solution...., but after re-thinking more sure this one is the right (and upstream acceptable) solution suitable for master and next...
Regards,
Peter
[1] https://patchwork.ozlabs.org/patch/1247577/
Quoted 128 lines
Regards,
Yann E. MORIN.Quoted 123 lines
../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: In member function 'bool
nv50_ir::GCRA::simplify()':
../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1348:19: error: expected unqualified-id before '(' tokenif (std::isinf(bestScore)) { ^Signed-off-by Peter Seiderer <[email protected]>---
Notes:
- as Romain Naour pointed out ([1], [2]) the existing check inc99_math.hfailes for uclibc (because it impersonate glibc-2.2), so enable the original isinf solution for uclibc instead of the custom work aroundChanges v1 -> v2:
- update patch to usedefined()around__UCLIBC__
- update uclibc impersonate version to 2.2 in notes and fix wording[1] http://lists.busybox.net/pipermail/buildroot/2020-March/275624.html
[2] http://lists.busybox.net/pipermail/buildroot/2020-March/275735.html--- ...rt-isinf-for-uclibc-based-toolchains.patch | 30 ++++++++++++ ...ers-nouveau-codegen-nv50_ir_ra.cpp-p.patch | 47 ------------------- 2 files changed, 30 insertions(+), 47 deletions(-) create mode 100644 package/mesa3d/0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch delete mode 100644 package/mesa3d/0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patchhttps://patchwork.ozlabs.org/patch/1247577/
package/mesa3d/0004-c99_math-import-isinf-for-uclibc-based-toolchains.patch new file +30
-0@@ -0,0 +1,30 @@ +From 705490763fcc052defa2d49a5c5200775cd12c54 Mon Sep 17 00:00:00 2001 +From: Peter Seiderer <[email protected]> +Date: Wed, 4 Mar 2020 08:27:24 +0100 +Subject: [PATCH] c99_math: import isinf for uclibc based toolchains + +The same fix as for glibc >= 2.23 is needed for uclibc +based toolchains. + +[Upstream suggested: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4042] +Signed-off-by: Peter Seiderer <[email protected]> +--- + include/c99_math.h | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/include/c99_math.h b/include/c99_math.h +index e906c26aa54..c8a73a2d2c7 100644 +--- a/include/c99_math.h ++++ b/include/c99_math.h +@@ -190,7 +190,7 @@ fpclassify(double x) + * undefines those functions, which in glibc 2.23, are defined as macros rather + * than functions as in glibc 2.22. + */ +-#if __cplusplus >= 201103L && (__GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 23)) ++#if __cplusplus >= 201103L && (__GLIBC__ > 2 || (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 23) || defined(__UCLIBC__)) + #include <cmath> + + using std::fpclassify; +-- +2.25.1 +package/mesa3d/0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch deleted +0
-48@@ -1,47 +0,0 @@ -From f2ce27b761eeca301a109856f20ed9a8eb06796d Mon Sep 17 00:00:00 2001 -From: Fabrice Fontaine <[email protected]> -Date: Sat, 28 Dec 2019 21:49:43 +0100 -Subject: [PATCH] src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: prefix - isinf with std:: - -Prefix isinf with std:: to fix the following build failure with gcc 4.9: - -../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp: In member function 'bool nv50_ir::GCRA::simplify()': -../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1350:29: error: 'isinf' was not declared in this scope - if (isinf(bestScore)) { - ^ -../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:1350:29: note: suggested alternative: -In file included from /data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/random:38:0, - from /data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/bits/stl_algo.h:66, - from /data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/algorithm:62, - from ../src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp:26: -/data/buildroot/buildroot-test/instance-0/output/host/opt/ext-toolchain/arm-buildroot-linux-uclibcgnueabi/include/c++/4.9.4/cmath:621:5: note: 'std::isinf' - isinf(_Tp __x) - ^ - -Fixes: - - http://autobuild.buildroot.org/results/599dc3478ed65d36fbf9f5e9625691dfa813b530 - -Signed-off-by: Fabrice Fontaine <[email protected]> -[Upstream status: -https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3227] ---- - src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp -index 6df2664da22..e81fb117213 100644 ---- a/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp -+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_ra.cpp -@@ -1345,7 +1345,7 @@ GCRA::simplify() - bestMaxReg = it->maxReg; - } - } -- if (isinf(bestScore)) { -+ if (std::isinf(bestScore)) { - ERROR("no viable spill candidates left\n"); - return false; - } --- -2.24.0 - --2.25.1
_______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot
PATCH v2 1/1 package/mesa3d: fix nouveau std::isinf related compile failure
Peter, All,
On 2020-03-04 22:11 +0100, Peter Seiderer spake thusly:
Quoted 12 lines · Yann E. MORIN
Quoted 11 lines
On 2020-03-04 12:00 +0100, Peter Seiderer spake thusly:
Quoted 10 lines
Activate already existing mesa3d solution for the isinf compile failure for uclibc based toolchains instead of using a custom workaround.
- remove 0005-src-gallium-drivers-nouveau-codegen-nv50_ir_ra.cpp-p.patch
- add 0004-c99_math-import-isinf-for-uclibc-based-toolchains.patchFixes:
http://autobuild.buildroot.net/results/cbefc5d4a4fefb674e596400fa1d2698cd89c5b3/
http://autobuild.buildroot.net/results/dc974da012f53fa4ed3be616f937b0afae423d66/
Quoted 8 lines
Quoted 3 lines
You've tagged this patch for next, but the two failures pccured on master.
Where should this patch be applied, then?The patch [1] was meant for master (minimal change, minimal risk) and this one (v1/v2) for next..., chronological later, first not sure about the solution...., but after re-thinking more sure this one is the right (and upstream acceptable) solution suitable for master and next...
Followign this discusion, I've applied it to master (rather than next). Thanks!
Regards,
Yann E. MORIN.
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------. | Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 561 099 427 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. | '------------------------------^-------^------------------^--------------------'