This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug middle-end/51994] [4.6/4.7 Regression] git-1.7.8.3 miscompiled due to negative bitpos from get_inner_reference
- From: "ubizjak at gmail dot com" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Wed, 25 Jan 2012 18:17:25 +0000
- Subject: [Bug middle-end/51994] [4.6/4.7 Regression] git-1.7.8.3 miscompiled due to negative bitpos from get_inner_reference
- Auto-submitted: auto-generated
- References: <bug-51994-4@http.gcc.gnu.org/bugzilla/>
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51994
Uros Bizjak <ubizjak at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #26462|0 |1
is obsolete| |
--- Comment #12 from Uros Bizjak <ubizjak at gmail dot com> 2012-01-25 18:17:25 UTC ---
Created attachment 26466
--> http://gcc.gnu.org/bugzilla/attachment.cgi?id=26466
WIP gcc-46 patch, includes other expmed.c store/extract bit functions
This patch fixes regression at gcc.dg/pr48335-8.c on x86_64-linux from previous
patch. However, some kind of unterminated loop is now triggered for this test
on x86_64 target:
(gdb) up
#2 0x00000000005b1a3d in extract_split_bit_field (op0=0x7ffff1340a80,
bitsize=16, bitpos=-24, unsignedp=1)
at ../../gcc-svn/branches/gcc-4_6-branch/gcc/expmed.c:1996
1996 word = operand_subword_force (op0, offset, GET_MODE (op0));
(gdb) up
#3 0x00000000005b1962 in extract_split_bit_field (op0=0x7ffff1340a80,
bitsize=16, bitpos=-24, unsignedp=1)
at ../../gcc-svn/branches/gcc-4_6-branch/gcc/expmed.c:2006
2006 part = extract_fixed_bit_field (word_mode, word,
(gdb) up
#4 0x00000000005b1962 in extract_split_bit_field (op0=0x7ffff1340a80,
bitsize=16, bitpos=-24, unsignedp=1)
at ../../gcc-svn/branches/gcc-4_6-branch/gcc/expmed.c:2006
2006 part = extract_fixed_bit_field (word_mode, word,
I was not able to fix it... definitelly for someone more experienced in this
area of the compiler.