MatCat.OpenSource/8sa1-gcc
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Comment to document possible problems with last change.

Browse Source 
* dwarf2out.c (output_call_frame_info): Comments on last change.

From-SVN: r23512
This commit is contained in:
Jim Wilson 1998-11-03 17:52:59 +00:00 committed by Jim Wilson
parent 8452b1d3c1
commit eef906d616
2 changed files with 12 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
gcc/ChangeLog
View File

@ -1,3 +1,7 @@
Tue Nov 3 17:51:36 1998 Jim Wilson <wilson@cygnus.com>
* dwarf2out.c (output_call_frame_info): Comments on last change.
Tue Nov 3 07:51:43 1998 Richard Earnshaw (rearnsha@arm.com) Tue Nov 3 07:51:43 1998 Richard Earnshaw (rearnsha@arm.com)
* arm.c (add_constant): When taking the address of an item in the * arm.c (add_constant): When taking the address of an item in the

8
gcc/dwarf2out.c
View File

@ -1829,6 +1829,14 @@ output_call_frame_info (for_eh)
fputc ('\n', asm_out_file); fputc ('\n', asm_out_file);
ASM_OUTPUT_LABEL (asm_out_file, l1); ASM_OUTPUT_LABEL (asm_out_file, l1);
/* ??? This always emits a 4 byte offset when for_eh is true, but it
emits a target dependent sized offset when for_eh is not true.
This inconsistency may confuse gdb. The only case where we need a
non-4 byte offset is for the Irix6 N64 ABI, so we may lose SGI
compatibility if we emit a 4 byte offset. We need a 4 byte offset
though in order to be compatible with the dwarf_fde struct in frame.c.
If the for_eh case is changed, then the struct in frame.c has
to be adjusted appropriately. */
if (for_eh) if (for_eh)
ASM_OUTPUT_DWARF_DELTA4 (asm_out_file, l1, "__FRAME_BEGIN__"); ASM_OUTPUT_DWARF_DELTA4 (asm_out_file, l1, "__FRAME_BEGIN__");
else else