- User Since
- Mar 8 2022, 3:45 AM (23 w, 2 d)
Wed, Jul 27
I think @xry111's suggestion to make this optimization tunable is reasonable, but it may need more work then. Perhaps this patch could be put on the back burner, we can always come to finish this later after the more essential things. Lots of downstream projects are blocked by availability of LLVM so it may be very worthwhile to shift priorities for now.
Tue, Jul 26
Mon, Jul 25
Thu, Jul 21
Is https://github.com/loongson/LoongArch-Documentation/pull/47 (using 0x1, 0x2, 0x3 for ILP32S/F/D and use EI_CLASS to distinguish LP64 and ILP32) dead?
Jul 19 2022
Jul 16 2022
I'm now feeling guilty because I've raise the suggestion w/o any benchmarking done... When I get some spare time I'll try to implement this for GCC and benchmark it.
Jul 15 2022
I don't have write access to rG, if you think these two changes are OK please commit it.
Remove unnecessary check for __loongarch__ before __loongarch_lp64.
Define __interception::real_vfork as a normal bss object, instead of common symbol.
nit: use one space instead of two after ..
Jul 13 2022
To other reviewers: we tested this optimization with an internal llvm version (llvm13) on 3A5000, and it shows that SPEC CPU2006 FP score increases 1% in average. 470.lbm score increases 8.9%. But we wonder why other architectures have not done so? Is there any potential issue?
Seems that currently we can only check the changes by GCC because LoongArch hasn’t been supported by clang.
Not sure whether this change could land. What do others think?
Jul 11 2022
Restore compiler check for sizeof(struct stat).
Use full diff
Use full diff
Jul 10 2022
Keep lexicographic order
Mark __interceptor::real_vfork with ASM_HIDDEN and use la.local for it.
Jul 9 2022
Remove unnecessary brackets.
Jul 8 2022
Use __loongarch_lp64 instead of __loongarch_grlen == 64 (as an ILP32 environment on LA64 CPU can also report __loongarch_grlen == 64).
Jul 6 2022
Jul 5 2022
Jul 4 2022
Jul 2 2022
Jul 1 2022
Jun 28 2022
I'll submit a patch to change GCC behavior later (after fix PR106096, which is blocking bootstrap).
Jun 25 2022
It seems I can't remove an inline comment...
Jun 24 2022
Jun 7 2022
Does anyone have any idea about "why the CI is failing"? It seems happening for this batch of LoongArch patches and also one MIPS patch I've reviewed today. I guess something is wrong in the CI system or LLVM codebase.
May 10 2022
Good to me, but as I've only contributed an one-line change to LLVM let's to wait for someone else :).
May 3 2022
It seems the CI failed with some tests TIMEOUT. Maybe we need to adjust the test timeout for the CI.
Apr 20 2022
Mar 9 2022
@dvyukov Thanks for the review. But I don't have a commit access to llvm-project.
Mar 8 2022
clang-format is doing strange things, if I removed those two spaces for indent it will then tell me to add them back...