[Bug lto/50568] New: [4.7 Regression] Massive LTO failures

[Bug lto/50568] New: [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

             Bug #: 50568
           Summary: [4.7 Regression] Massive LTO failures
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: lto
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: hjl.tools@gmail.com


On Linux/x86, revision 179350 has massive LTO failures:

FAIL: c-c++-common/guality/pr43141.c  -O2 -flto  (internal compiler error)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto  (internal compiler error)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto  (test for excess errors)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto  (test for excess errors)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto -flto-partition=none  (internal
compiler error)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto -flto-partition=none  (internal
compiler error)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto -flto-partition=none  (test for
excess errors)
FAIL: c-c++-common/guality/pr43141.c  -O2 -flto -flto-partition=none  (test for
excess errors)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-alias-1.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-sign-add.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-add.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-div.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-mul.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mixed-sub.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto  (test for
excess errors)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mul-minus-one.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto  (internal
compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto
-flto-partition=none  (internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mul-one.c  -O2 -flto
-flto-partition=none  (test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-sign-mul.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto  (internal compiler
error)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto  (test for excess
errors)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto -flto-partition=none 
(internal compiler error)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/complex-sign-sub.c  -O2 -flto -flto-partition=none 
(test for excess errors)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto  (internal compiler error)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto  (internal compiler error)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto  (test for excess errors)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto  (test for excess errors)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto -flto-partition=none  (internal
compiler error)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto -flto-partition=none  (internal
compiler error)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto -flto-partition=none  (test for
excess errors)
FAIL: c-c++-common/torture/pr42834.c  -O2 -flto -flto-partition=none  (test for
excess errors)
FAIL: g++.dg/ipa/pr46984.C (internal compiler error)
FAIL: g++.dg/ipa/pr46984.C (test for excess errors)
...

Revision 179346 is OK.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andi-gcc at firstfloor dot
                   |                            |org

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 15:50:30 UTC ---
It may only happen on Linux/ia32.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #2 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 15:58:26 UTC ---
Looking...

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 15:59:54 UTC ---
I got

lto1: internal compiler error: resolution sub id ffffffff not in object file^M
Please submit a full bug report,^M
with preprocessed source if appropriate.^M
See <http://gcc.gnu.org/bugs.html> for instructions.^M

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.7.0

--- Comment #4 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 16:02:04 UTC ---
It is caused by revision 179348:

http://gcc.gnu.org/ml/gcc-cvs/2011-09/msg00967.html

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 17:56:25 UTC ---
The problem is in

Breakpoint 2, process_symtab (data=0xffffccc0, 
    name=0x82041fe ".gnu.lto_.symtab.f1d7150d3f9de9cb", offset=1325, length=56)
    at /export/gnu/import/git/gcc/lto-plugin/lto-plugin.c:813
813      secdata = xmalloc (length);
(gdb) p s
$1 = 0x820420e ".f1d7150d3f9de9cb"
(gdb) p obj->out->id
$2 = 4294967295
(gdb) p/x obj->out->id
$3 = 0xffffffff
(gdb) 

We got 32bit overflow with 64bit random seed.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #6 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 18:03:21 UTC ---
I don't see the problem on a 64bit bootstrap-lto.

I guess i must have written some 32bit unsafe code.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #7 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:11:50 UTC ---
(In reply to comment #6)
> I don't see the problem on a 64bit bootstrap-lto.
> 
> I guess i must have written some 32bit unsafe code.

We can't use 64bit random seed when LTO expects 32bit value.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #8 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:16:03 UTC ---
Created attachment 25380
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25380
A patch

This patch works for me.  But I don't think it is correct.
We need a way to specify HOST_WIDE_INT for lto plugin.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #9 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 18:17:08 UTC ---
Created attachment 25381
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25381
Use long long in lto-plugin

Can you please test this patch?

Thanks.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #10 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 18:19:02 UTC ---
I did the same patch (with long long)

I think using long long here is ok because lto-plugin only builds
on modern and non weird hosts and they should all have long long 
anyways.

uint64_t is probably fine too.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #11 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:19:30 UTC ---
HOST_WIDE_INT is needed for gcc, libcpp and plug-in.  We should
have a central host-wide-int.m4 to define all HOST_WIDE_INT related
macros.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #12 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:21:17 UTC ---
(In reply to comment #9)
> Created attachment 25381 [details]
> Use long long in lto-plugin
> 
> Can you please test this patch?
> 

It won't work since you also need to update lto-plugin/lto-plugin.c
for 64bit id.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #13 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:26:06 UTC ---
(In reply to comment #8)
> Created attachment 25380 [details]
> A patch
> 
> This patch works for me.  But I don't think it is correct.
> We need a way to specify HOST_WIDE_INT for lto plugin.

That won't work since splay-tree doesn't support 64bit key on
32bit hosts.  We can't use 64bit id for LTO.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #14 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 18:27:11 UTC ---
But that's what I did?

% diffstat plugin-fix 
 lto-plugin.c |   11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)



I don't see why long long cannot be used on the platforms supporting
plugins (windows, darwin, linux)

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #15 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 18:28:18 UTC ---
Hmm good point.

Maybe the splay tree can be fixed. Otherwise have to use 32bit ids on 32bit,
but then the risk of collisions is higher again.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Depends on|                            |34835

--- Comment #16 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:29:19 UTC ---
(In reply to comment #15)
> Hmm good point.
> 
> Maybe the splay tree can be fixed. Otherwise have to use 32bit ids on 32bit,
> but then the risk of collisions is higher again.

See PR 34835.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #17 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 18:34:38 UTC ---
Also I don't believe it is 100% safe to use %x for printf/scanf
on 64bit integer even on 64bit hosts. I think 64bit random seed
change should be reverted for now.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #18 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 20:36:50 UTC ---
Created attachment 25384
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25384
fix + splay tree

I have some unrelated trouble with a 32bit bootstrap currently.

This patch should fix all the problems with the splay tree, by allocating
the key separately.

Can you give it a test please? Thanks

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #19 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 20:44:44 UTC ---
(In reply to comment #18)
> Created attachment 25384 [details]
> fix + splay tree
> 
> I have some unrelated trouble with a 32bit bootstrap currently.
> 
> This patch should fix all the problems with the splay tree, by allocating
> the key separately.
> 

2 comments:

1. We should use long long consistently in both lto-plugin
and lto. We can add a check to enable LTO only if long long is
supported on host.
2.  Can we avoid indirect penalty on 64bit hosts?

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #21 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 22:20:02 UTC ---
(In reply to comment #20)
> HOST_WIDE_INT may not be 64bit on 32bit host.
> But long long is 64bit.  I don't think it is
> correct to use HOST_WIDE_INT in lto.

It may not be a problem since lto-plugin.c can use
long long for symbol ID encoded in section name by
lto.c even if HOST_WIDE_INT is 32bit.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #20 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 22:14:28 UTC ---
HOST_WIDE_INT may not be 64bit on 32bit host.
But long long is 64bit.  I don't think it is
correct to use HOST_WIDE_INT in lto.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #22 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 22:23:53 UTC ---
(In reply to comment #18)
> Created attachment 25384 [details]
> fix + splay tree
> 
> I have some unrelated trouble with a 32bit bootstrap currently.
> 
> This patch should fix all the problems with the splay tree, by allocating
> the key separately.
> 
> Can you give it a test please? Thanks

Doesn't work:

gdb) bt
#0  0x080def1f in lto_splay_compare_ids (a=484680112, b=153199512)
    at /export/gnu/import/git/gcc/gcc/lto/lto.c:1109
#1  0x08d26ee8 in splay_tree_splay (sp=0x921a240, key=484680112)
    at /export/gnu/import/git/gcc/libiberty/splay-tree.c:149
#2  0x08d274ef in splay_tree_lookup (sp=0x921a240, key=484680112)
    at /export/gnu/import/git/gcc/libiberty/splay-tree.c:456
#3  0x080dec01 in create_subid_section_table (slot=0x921a168, data=0x921a240)
    at /export/gnu/import/git/gcc/gcc/lto/lto.c:1022
#4  0x08d21882 in htab_traverse_noresize (htab=0x919bf50, 
    callback=0x80debb4 <create_subid_section_table>, info=0x921a240)
    at /export/gnu/import/git/gcc/libiberty/hashtab.c:784
#5  0x08d218ea in htab_traverse (htab=0x919bf50, 
    callback=0x80debb4 <create_subid_section_table>, info=0x921a240)
    at /export/gnu/import/git/gcc/libiberty/hashtab.c:800
#6  0x080def8a in lto_file_read (file=0x919c9f0, resolution_file=0x9219fe0, 
    count=0xffffd08c) at /export/gnu/import/git/gcc/gcc/lto/lto.c:1132
#7  0x080e5a43 in read_cgraph_and_symbols (nfiles=1, fnames=0x91a8060)
    at /export/gnu/import/git/gcc/gcc/lto/lto.c:2496
#8  0x080e6196 in lto_main () at /export/gnu/import/git/gcc/gcc/lto/lto.c:2809
#9  0x085dadfb in compile_file ()
    at /export/gnu/import/git/gcc/gcc/toplev.c:565
#10 0x085dcf2e in do_compile () at /export/gnu/import/git/gcc/gcc/toplev.c:1925
#11 0x085dd0b0 in toplev_main (argc=16, argv=0x9181450)
---Type <return> to continue, or q <return> to quit---
    at /export/gnu/import/git/gcc/gcc/toplev.c:2001
#12 0x080e8be7 in main (argc=16, argv=0xffffd224)
    at /export/gnu/import/git/gcc/gcc/main.c:36
(gdb)

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #23 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 22:26:34 UTC ---
(In reply to comment #9)
> Created attachment 25381 [details]
> Use long long in lto-plugin
> 
> Can you please test this patch?
> 

You missed:

   /* Find hash table of sub module id */
-  nd = splay_tree_lookup (file_ids, id);
+  nd = splay_tree_lookup (file_ids, (splay_tree_key)&id);

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #24 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 23:06:35 UTC ---
Thanks. Does it work with this change?

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #25 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 23:16:20 UTC ---
Created attachment 25386
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25386
A better patch

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #26 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 23:19:49 UTC ---
(In reply to comment #24)
> Thanks. Does it work with this change?

I posted a different patch to avoid indirect reference on
64bit host and avoid overflow in

static int lto_splay_compare_ids (splay_tree_key a, splay_tree_key b)
{
   unsigned HOST_WIDE_INT *ai = (unsigned HOST_WIDE_INT *)a;
   unsigned HOST_WIDE_INT *bi = (unsigned HOST_WIDE_INT *)b;

   return *ai - *bi;
}

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #27 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 23:21:12 UTC ---
Hmm is that just for efficiency or did you fix another bug?

(not worrying about efficiency too much because this tree has only
one entry per input file)

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #28 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 23:22:17 UTC ---
I don't understand which overflow you refer to. Can you please clarify?

afaik a - b is the standard way to write these comparison functions.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #29 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-29 23:31:13 UTC ---
(In reply to comment #28)
> I don't understand which overflow you refer to. Can you please clarify?
> 
> afaik a - b is the standard way to write these comparison functions.

lto_splay_compare_ids returns int. *ai - *bi may be 0xfffffff10000000
and lto_splay_compare_ids returns > 0 instead of < 0.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #30 from Andi Kleen <andi-gcc at firstfloor dot org> 2011-09-29 23:33:52 UTC ---
Okay. Can you post the patch then?

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                URL|                            |http://gcc.gnu.org/ml/gcc-p
                   |                            |atches/2011-09/msg01977.htm
                   |                            |l

--- Comment #31 from H.J. Lu <hjl.tools at gmail dot com> 2011-09-30 00:57:59 UTC ---
A patch posted at

http://gcc.gnu.org/ml/gcc-patches/2011-09/msg01977.html

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

--- Comment #32 from hjl at gcc dot gnu.org <hjl at gcc dot gnu.org> 2011-09-30 15:48:56 UTC ---
Author: hjl
Date: Fri Sep 30 15:48:51 2011
New Revision: 179395

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=179395
Log:
Use 64bit integer for LTO symbol ID.

gcc/lto

2011-09-30  H.J. Lu  <hongjiu.lu@intel.com>
        Andi Kleen  <ak@linux.intel.com>

    PR lto/50568
    * lto.c (lto_splay_tree_delete_id): New.
    (lto_splay_tree_compare_ids): Likewise.
    (lto_splay_tree_lookup): Likewise.
    (lto_splay_tree_id_equal_p): Likewise.
    (lto_splay_tree_insert): Likewise.
    (lto_splay_tree_new): Likewise.
    (lto_resolution_read): Change id to unsigned HOST_WIDE_INT.
    Use lto_splay_tree_id_equal_p and lto_splay_tree_lookup.
    (create_subid_section_table): Use lto_splay_tree_lookup and
    lto_splay_tree_insert.
    (lto_file_read): Use lto_splay_tree_new.

lto-plugin/

2011-09-30  H.J. Lu  <hongjiu.lu@intel.com>
        Andi Kleen  <ak@linux.intel.com>

    PR lto/50568
    * lto-plugin.c (sym_aux): Change id to unsigned long long.
    (plugin_symtab): Likewise.
    (dump_symtab): Likewise.
    (resolve_conflicts): Likewise.
    (process_symtab): Likewise.

Modified:
    trunk/gcc/lto/ChangeLog
    trunk/gcc/lto/lto.c
    trunk/lto-plugin/ChangeLog
    trunk/lto-plugin/lto-plugin.c

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568

Richard Guenther <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |FIXED

--- Comment #33 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-10-10 15:14:07 UTC ---
Fixed.

[Bug lto/50568] [4.7 Regression] Massive LTO failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=50568
Bug 50568 depends on bug 34835, which changed state.

Bug 34835 Summary: splay-tree doesn't support 64bit value on 32bit host
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=34835

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |MOVED