Summary: | [3.4/4.0 Regression] Option -mms-bitfields support on GCC 3.4 is not conformant to MS layout | ||
---|---|---|---|
Product: | gcc | Reporter: | Pascal Obry <p.obry> |
Component: | target | Assignee: | Not yet assigned to anyone <unassigned> |
Status: | RESOLVED FIXED | ||
Severity: | critical | CC: | dannysmith, dj, gcc-bugs |
Priority: | P2 | Keywords: | wrong-code |
Version: | 3.4.0 | ||
Target Milestone: | 3.4.1 | ||
Host: | pentium-mingw32msv | Target: | |
Build: | pentium-mingw32msv | Known to work: | |
Known to fail: | Last reconfirmed: | 2004-05-26 11:39:53 |
Description
Pascal Obry
2004-05-02 17:41:55 UTC
gcc-dg/bf-ms-layout.c has been failing since at least 4 Sept 2003: http://gcc.gnu.org/ml/gcc-testresults/2003-09/msg00126.html Danny This delta was part of large patchset committed by DJ Delorie on 2003-09-03. It is not recorded in ChangeLog. Reverting it restores correct ms-bitfield-layout. =================================================================== RCS file: /cvs/gcc/gcc/gcc/stor-layout.c,v retrieving revision 1.166 retrieving revision 1.167 diff -u -r1.166 -r1.167 --- gcc/gcc/stor-layout.c 2003/08/20 12:24:18 1.166 +++ gcc/gcc/stor-layout.c 2003/09/04 03:17:50 1.167 @@ -1093,6 +1093,7 @@ rli->prev_field = NULL; } + rli->offset_align = tree_low_cst (TYPE_SIZE (type), 0); normalize_rli (rli); } Subject: Re: Option -mms-bitfields support on GCC 3.4 is not conformant to MS layout
> This delta was part of large patchset committed by DJ Delorie on 2003-09-03.
>
> It is not recorded in ChangeLog.
Accident :-P
Subject: Re: Option -mms-bitfields support on GCC 3.4 is not conformant to MS layout dannysmith at users dot sourceforge dot net writes: > > ------- Additional Comments From dannysmith at users dot sourceforge dot net 2004-05-03 01:35 ------- > This delta was part of large patchset committed by DJ Delorie on 2003-09-03. > > It is not recorded in ChangeLog. > > Reverting it restores correct ms-bitfield-layout. I confirm that reverting this patch does fix the problem on my side. Do you think this will be part of GCC 3.4.1 ? Thanks, Pascal. -- --|------------------------------------------------------ --| Pascal Obry Team-Ada Member --| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE --|------------------------------------------------------ --| http://www.obry.org --| "The best way to travel is by means of imagination" --| --| gpg --keyserver wwwkeys.pgp.net --recv-key C1082595 Confirmed. DJ -- Would you please revert the erroneous part of the patch ASAP? -- Mark Subject: Re: [3.4/3.5 Regression] Option -mms-bitfields support on GCC 3.4 is not conformant to MS layout > DJ -- > > Would you please revert the erroneous part of the patch ASAP? > > -- Mark > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15250 Done, for head and 3.4 2004-05-28 DJ Delorie <dj@redhat.com> * stor-layout.c (place_field): Revert erroneous commit. Index: stor-layout.c =================================================================== RCS file: /cvs/gcc/gcc/gcc/stor-layout.c,v retrieving revision 1.190 diff -p -C2 -r1.190 stor-layout.c *** stor-layout.c 19 May 2004 17:53:45 -0000 1.190 --- stor-layout.c 28 May 2004 23:11:34 -0000 *************** place_field (record_layout_info rli, tre *** 1128,1132 **** } - rli->offset_align = tree_low_cst (TYPE_SIZE (type), 0); normalize_rli (rli); } --- 1128,1131 ---- Okay only a 3.5.0 regression now. Fixed. |