Just a comment to point that this patch will help building the RISCV Architecture test-suite with Clang
where test macros for JAL and branch contain instructions like inst reg1, reg2, label+offset.

I found an issue with the IROutliner while testing this patch which I don't easily catch, when pruning icompatible regions there is an assert violation at llvm/include/llvm/ADT/ilist_iterator.h:138
The problem can be reproduce with this reduce testcase:

opt -S -iroutliner ~/test.ll -o -

test.ll:

@a = global i8* null

Apart from the previous comments, this patch LGTM and fixes a correctness issue observed on Spec, Thanks for that

LGTM Thanks.

LGTM, I confirm that it fixes the alignement issue observed on Thumbv7 bots

Thank you Yvan!
Potential fix: https://reviews.llvm.org/D104402
The new DCHECK caught the alignment issue on x86/64, I assume this would work for armhf as well.

this is due to a misaligned data access, in Thumb the code generated for unmapTestOnly
contains a Vector Store instruction vst1.64 {d16-d17}, [r0, :128] which needs a 16-bytes
alignement, but the adress of TSDs (which is in r0) is not.

This change introduced a regression on ARMv7 Thumb bots which wasn't noticed due to other issues.

In D103167#2802543, @chill wrote:

LGTM, just if you intend to reuse the summary as a commit message "... since MVE instructions are *not* outlined...".

@efriedma, @chill is this patch ok for you ? (it'd be great to have the issue fixed into 12.0.1 release)

Hi Vitaly,

Hi and Thanks for working on this,

LGTM as well

I've committed this and it looks like the bot is back to green :) https://lab.llvm.org/buildbot/#/builders/65/builds/1400. Sorry for the delayed fix

Sorry I'm bit lost in the various patches proposed to fix the issue introduced by https://reviews.llvm.org/D97785
My understanding is that this is missing one to fix our Windows on ARM bots (broken for more than 2 weeks now)
So it'd be great to have it applied

Hi Abhina,

Hmm, there is tools/llvm-nm/debug-syms.test which tests ARM --special-syms and --debug-syms, notice that with this change --special-syms is no more needed to dump mapping-symbols when --debug-syms is used, which is now in sync with gnu binutils nm. I don't see an AArch64 test though

This new version should just change llvm-nm behavior for ELF ARM and AArch64 targets to handle SF_FormatSpecific symbols and doesn't change the processing of COFF or debug symbols

Sure, because in dumpSymbolNamesFromObject line1815 these symbols are skipped if they are not debug ones, thus not printed even if --special-syms flag is used.

Thanks for the reviews,

I'm on a D05 box (Cortex-A72) and I should have a remote access to WoA surface laptop... I'll check

Hi Sam,

I'm not experienced enough in the machine scheduler to approve it, but the description of the problem and the proposed solution make sense to me.

It makes sense, thanks.

Hi Sam,

@lenary it'd be great if you can commit your fix today, the bots are red for long time

this commit broke ARM bots, because their data model is ILP32, thus it is not safe to use a long format specifier to print a 64 bits type.

Hi Petr,

All patches are now committed.

Hi Sam,

LGTM

ping

Approved now that I've properly tested it.

One small comment otherwise LGTM

ping

ping

Hi Andrew, the new version LGTM

Hi Sam,

Ping

I think that the previous comments were addressed. LGTM

Just a small comment, otherwise LGTM, thanks

Hi David,

LGTM

LGTM, given it is needed to bring ARM bots back online

The change makes sense to me and will fix ARM broken bots, I'd say that it can be accepted if there is no objection within a day

LGTM

I'd check if there are some calls of modifiesRegister that pass nullptr, the patch LGTM otherwise

(Sorry for the delay, I took some vacations)

Thanks for raising this issue, I've two inlined comments

Hi Louis,