Re: Memory alignment not respected in thread-local objects

[Joerg Sonnenberger <joerg%bec.de@localhost>]

On Thu, Dec 20, 2018 at 10:21:30PM +0100, Michał Górny wrote:
> On Thu, 2018-12-20 at 20:56 +0100, Joerg Sonnenberger wrote:
> > On Thu, Dec 20, 2018 at 08:18:57PM +0100, Michał Górny wrote:
> > > My today's problem was to evaluate crashes in compiler-rt's xray tests. 
> > > After some debugging, I've came to the conclusion that the primary cause
> > > is that TLS allocated for aligned objects isn't actually aligned.
> > 
> > That's correct, TLS alignment is currrently mostly ignored. I don't have
> > a short term plan for fixing that.
> > 
> 
> How about forcing a larger default alignment?  64 bytes would be
> sufficient for our compiler-rt, and I don't think we need to be sparing
> with the address space here.

No easy approach for that either.

Joerg