[llvm] r229841 - Reverting r229831 due to multiple ARM/PPC/MIPS build-bot failures.

[Kuperstein, Michael M]

We can, but I'll need someone on the arm side to help me debug. :-)

Aside from the MIPS failures, I was also getting self-host failures on arm - specifically, stage 2 crashes (or maybe out-of-memory, don't remember) when trying to compile the new tablegen-generated files. 
It seems like this doesn't happen on x86, and I'm not sure whether it's really a miscompile of the stage 1 clang caused by the change, or whether it just uncovers a front-end bug.

I can rebase the patch and try to commit again, I'll need someone on the ARM side to help me out with the debugging.
As an alternative, if you could try to reproduce the crash with r229841 or r233055 - e.g. try to self-host one of them, or build a clean top-of-trunk clang on a configuration similar to the armv7-a15-neon buildbot, and try to build r229841/r233055 with it, to see whether we really have a frontend issue, that would be great.

Thanks,
   Michael

-----Original Message-----
From: Richard Barton [mailto:richard.barton at arm.com] 
Sent: Monday, May 11, 2015 19:22
To: Kuperstein, Michael M
Cc: llvm-commits at cs.uiuc.edu
Subject: RE: [llvm] r229841 - Reverting r229831 due to multiple ARM/PPC/MIPS build-bot failures.

Hi Michael

Now that Daniel has upgraded the gcc toolchain on the MIPS buildbot, I was wondering if there any reason why we cannot try and re-apply the patch to introduce the FeatureBitset to MC?

Thanks
Rich



---------------------------------------------------------------------
Intel Israel (74) Limited

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.