public inbox for
 help / color / mirror / Atom feed
From: Richard Biener <>
To: Jeff Law <>
Cc:, Jan Hubicka <>
Subject: Re: [PATCH] Improve DSE to handle stores before __builtin_unreachable ()
Date: Wed, 21 Jun 2023 06:41:38 +0000 (UTC)	[thread overview]
Message-ID: <nycvar.YFH.7.77.849.2306210620560.4723@jbgna.fhfr.qr> (raw)
In-Reply-To: <>

On Tue, 20 Jun 2023, Jeff Law wrote:

> On 6/20/23 00:59, Richard Biener via Gcc-patches wrote:
> > DSE isn't good at identifying program points that end lifetime
> > of variables that are not associated with virtual operands.  But
> > at least for those that end basic-blocks we can handle the simple
> > case where this ending is in the same basic-block as the definition
> > we want to elide.  That should catch quite some common cases already.
> > 
> > Bootstrapped and tested on x86_64-unknown-linux-gnu.
> > 
> > As you can see from the testcase I had to adjust this possibly can
> > lead to more severe issues when one forgets a return (the C++ frontend
> > places builtin_unreachable () there).  I'm still planning to push
> > this improvement unless I hear objections.
> > 
> > Thanks,
> > Richard.
> > 
> >  * (dse_classify_store): When we found
> >  no defs and the basic-block with the original definition
> >  ends in __builtin_unreachable[_trap] the store is dead.
> > 
> >  * gcc.dg/tree-ssa/ssa-dse-47.c: New testcase.
> >  * c-c++-common/asan/pr106558.c: Avoid undefined behavior
> >  due to missing return.
> I thought during the introduction of erroneous path isolation that we
> concluded stores, calls and such had observable side effects that must be
> preserved, even when we hit a block that leads to __builtin_unreachable.

Indeed, I remember we repeatedly hit this in the past.  But 
double-checking I see that we instrument

  if (x)
    *(int *)0 = 0;


  <bb 2> [local count: 1073741824]:
  if (x_2(D) != 0)
    goto <bb 3>; [50.00%]
    goto <bb 4>; [50.00%]

  <bb 3> [local count: 536870913]:
  MEM[(int *)0B] ={v} 0;
  __builtin_trap ();

path isolation doesn't seem to use __builtin_unreachable ().  I did
not add __builtin_trap () as possible sink (but I did want to treat
__builtin_unreachable () and __builtin_unreachable_trap () the same
way).  The pass also marks the offending store as volatile.

We do have testsuite coverage that this happens (dump-scanning,
not runtime it seems).

So yes, I think preserving the original trap kind (if there is any)
is important and it still seems to work.  I don't remember whether
we have any test coverage for that though.  I'll also note that
__builtin_trap () has virtual operands (def and use) while
__builtin_unreachable[_trap] () are 'const'.  Honza correctly
says they should probably be novops instead of 'const' preserving
the fact that they have side-effects.

> Don't get me wrong, I'm all for removing the memory references if it's safe to
> do so.

I think it's desirable for assertions.  Since we elide plain
__builtin_unreachable () and fall thru whereever it leads that
shouldn't be an issue.

If I manually add a __builtin_unreachable () to the above case
I see the *(int *)0 = 0; store DSEd.  Maybe we should avoid
removing stores that might trap here?  POSIX wise such a trap
could be a way to jump out of the path leading to unreachable ()
via siglongjmp ...


  reply	other threads:[~2023-06-21  6:41 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <>
2023-06-20 13:27 ` Jeff Law
2023-06-21  6:41   ` Richard Biener [this message]
2023-06-21  9:55     ` Jan Hubicka
2023-06-21 14:04     ` Jeff Law
2023-06-22  6:31       ` Richard Biener
2023-06-22 13:36         ` Jeff Law
2023-06-22 13:42           ` Jan Hubicka
2023-06-24 14:24             ` Jeff Law
2023-06-25 16:33               ` Jan Hubicka
2023-06-26 17:21   ` Jan Hubicka
2023-06-26 22:37     ` Jeff Law
2023-06-20  6:59 Richard Biener

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=nycvar.YFH.7.77.849.2306210620560.4723@jbgna.fhfr.qr \ \ \ \ \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).