-
Thu May 07 2020 Qing Zhao <qing.zhao@oracle.com> 8.2.0-12
- Replace the previous 8 patches in version 11 with 19 patches from upstream
gcc 8 to support -moutline-atomics for aarch64
moutline-1-add-early-clobber.patch
moutline-2-simplify-lse-cas.patch
moutline-3-imp-cas-gen.patch
moutline-4-imp-swp-gen.patch
moutline-5-imp-atomic-op-lse.patch
moutline-6-remove-early-clobber.patch
moutline-7-extend-R-for-intreg.patch
moutline-8-imp-TImode-compare-and-swap.patch
moutline-9-fix-shrinkwrapping-interactions.patch
moutline-10-tidy-split-compare-and-swap.patch
moutline-11-out-of-line-for-lse-atomics.patch
moutline-12-visibility-to-libfunc.patch
moutline-13-imp-moutline-atomics.patch
moutline-14-fix-store-exclusive-lse.patch
moutline-15-configure-for-sys-auxv-libgcc-lse.patch
moutline-16-fix-compare-and-swphi.patch
moutline-17-fix-bootstrap-with-old-binutils.patch
moutline-18-fix-ice-dueto-gen-compare-reg.patch
moutline-19-re-pr90724.patch
-
Wed Apr 29 2020 Qing Zhao <qing.zhao@oracle.com> 8.2.0-11
- Add 8 patches from gcc9,10 to support -moutline-atomics for aarch64
gcc9-early-clobber-aarch64-store.patch
gcc9-gen-helper-substituting-iterat.patch
gcc9-imp-swp-generation.patch
gcc9-imp-atomic-op-lse.patch
gcc9-simplify-lse-cas.patch
gcc9-improve-cas.patch
gcc9-visibility-to-libfunc.patch
gcc10-imp-moutline-atomics.patch
-
Mon Jan 20 2020 Qing Zhao <qing.zhao@oracle.com> 8.2.0-10
- Add 5 patches from gcc9 to support Arm Ares and Neoverse-N1 for Aarch64
gcc9-add-mtune-support-for-arm-ares.patch
gcc9-add-vec-reverse.patch
gcc9-multiple-changes-align.patch
gcc9-initial-mcpu-ares-tuning.patch
gcc9-add-support-for-neoverse-n1.patch
-
Mon Nov 25 2019 Qing Zhao <qing.zhao@oracle.com> 8.2.0-9
- Fix Orabug 29968294 - Heap corruption with fprofile-dir=%p prevents
profiling parallel processes, needed for RDBMS:
Add patch to fix PR86057 from Gcc9:
gcc9-pr86057.patch
- Fix Orabug 30044244 - Profile directory concatenated with object file path
This is the same bug as gcc upstream PR91971:
gcc9-pr85759.patch
gcc10-pr91971.patch
-
Thu Apr 04 2019 Qing Zhao <qing.zhao@oracle.com> 8.2.0-8
- Fix Orabug 29599147 - Need -fprofile-dir=%q{VAR} backported to gcc8
This is the similar GCC PR47618, add the fix from Gcc9:
gcc9-pr47618.patch
- Fix Orabug 29457270 - With live patching, still need to obey "inline"
directive
This is the similar GCC PR89730, add the fix from Gcc9:
gcc9-pr89730.patch
-
Tue Mar 12 2019 Qing Zhao <qing.zhao@oracle.com> 8.2.0-7
- Fix Orabug 29178534 - SEGV in rdbms with gcc8 running lrgsrgdbcon1
lrgsrgdbcon2 lrgsrgdbcone
This is the similar GCC PR89595, add the fix from Gcc9:
gcc9-pr89595.patch
also fixes GCC PR PR89655 PR89656, add the fix from Gcc9:
gcc9-pr89655-pr89656.patch
-
Fri Jan 04 2019 Qing Zhao <qing.zhao@oracle.com> 8.2.0-6
- Add -fopt-info-inline support from Gcc9:
gcc9-opt-info-inline.patch
-
Fri Dec 14 2018 Qing Zhao <qing.zhao@oracle.com> 8.2.0-5
- Add -flive-patching support from Gcc9.
gcc9-fipa-reference-addressable.patch
gcc9-fipa-stack-alignment.patch
gcc9-add-fomit-frame-pointer-to-test.patch
gcc9-flive-patching.patch
gcc9-ipa-stack-alignment-386-test.patch
-
Thu Oct 18 2018 Qing Zhao <qing.zhao@oracle.com> 8.2.0-4
- replace the previous 17 ampere patches from
https://git.theobroma-systems.com/ampere-computing/gcc.git/log/?h=gcc-8_2_0-amp2-branch
with new 17 ampere patches from
https://git.theobroma-systems.com/ampere-computing/gcc.git/log/?h=gcc-8_2_0-amp3-branch
-
Thu Oct 04 2018 Qing Zhao <qing.zhao@oracle.com> 8.2.0-3
- Backport the following ampere patches from
https://git.theobroma-systems.com/ampere-computing/gcc.git/log/?h=gcc-8_2_0-amp2-branch:
d735f3ae4712f66362326d179b4d7e9332c79677: Revert "2017-10-24 Richard Biener
271e2811e59c0c77fc022fa86a7030f20b4cac8e: Correct the maximum shift amount for shifted
0512749950d927de3dd695f2f2aacdfd30cf32fd: Add CPU support for Ampere Computing's eMAG.
c8b87078f9e0714cb9cab602e12a18ceb12df05a: eMAG/Xgene: Procedural cost-model for X-Gene
1c35e4428832abc9ac9af16963067dbfe45b895b: Updating cost table for xgene1.
ceec28063f4654abb2d2a832e1ae72805fd3ae9e: Add Xgene1 prefetch tunings.
682c651cc55ec6899d069baf4567d2480cfa4d0a: Fix in xgene1_addrcost_table.
6c2f97f9803e0ae2618259b1cc103e1f25d9a3bb: aarch64: X-Gene: Adapt tuning struct for GCC 8.
cea6788434efe6dc9f647a10a756b61d82775814: tree-ssa-list-find-pipeline: Add pipelining loads
950779013ba24a3ca1fec5744b902f2f1bd25d48: uncse: Added pass to undo common subexpression
4f33e9bb7cedaadedf894d8da7a815ec712cc753: loop-prefetcher: Adapt defaults for X-Gene cores.
4aa975a6b380d17e6df5ede3a9ea5ef0d250c646: tree-ssa-cpp: Don't crash on SSA names without
c0238eaf2efdb6855bcb1b929d9c88c03e06234e: Introduce new command line flag -funroll-more.
516b7d1cd1015e45a0359aacfb3faa2a044e65fd: tree-vect-slp: Enable vectorization when
-fvectorize-more is given.
b67a04233bfc1f711b72495fdd09391aeaf7af82: cfgloopmanip: Allow forced creation of loop preheaders.
1a7503cd23d1084ebc6bd7f4892f0c2c5827e0ca: noloopalias: Add new pass to optimise loops.