[Bug gas/32952] sframe: incorrect handling of .cfi_undefined in gas

2025-05-29 Thread indu.bhagat at oracle dot com
https://sourceware.org/bugzilla/show_bug.cgi?id=32952

Indu Bhagat  changed:

   What|Removed |Added

 Resolution|--- |FIXED
 Status|ASSIGNED|RESOLVED

--- Comment #3 from Indu Bhagat  ---
Fixed for 2.45

-- 
You are receiving this mail because:
You are on the CC list for the bug.


[Bug gas/32953] sframe: incorrect handling of .cfi_same_value in gas

2025-05-29 Thread indu.bhagat at oracle dot com
https://sourceware.org/bugzilla/show_bug.cgi?id=32953

Indu Bhagat  changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |FIXED

--- Comment #4 from Indu Bhagat  ---
Fixed on master

-- 
You are receiving this mail because:
You are on the CC list for the bug.


[Bug gas/33030] New: [Z80] [PATCH] accept leading zeros on dollar local labels when using -sdcc

2025-05-29 Thread aargri at gmail dot com
https://sourceware.org/bugzilla/show_bug.cgi?id=33030

Bug ID: 33030
   Summary: [Z80] [PATCH] accept leading zeros on dollar local
labels when using -sdcc
   Product: binutils
   Version: 2.45 (HEAD)
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: gas
  Assignee: unassigned at sourceware dot org
  Reporter: aargri at gmail dot com
  Target Milestone: ---

Created attachment 16109
  --> https://sourceware.org/bugzilla/attachment.cgi?id=16109&action=edit
gas: accept leading zeros on dollar local labels for z80 sdcc compat

SDCC uses zero-padded 5 digit numbers as dollar local labels in its assembly
output. GAS used to accept this, but this commit removed support for such
labels:

https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=226749d5a6ff0d5c607d6428d6c81e1e7e7a994b

As a result, GAS can no longer compile SDCC's output, even with the -sdcc flag.

I've attached a patch to restore SDCC compatibility when using the -sdcc flag,
by recognizing dollar local label definitions at the beginning of the line and
replacing the leading zeros with spaces. This works fine, although it leaves
the references to these labels inside instructions unchanged.

My patch also restores the SDCC test case to use 5-digit labels. It seems like
it was modified to force it to pass when this change was first made. (?!)

I think this is a reasonable fix, but this is a complicated code base that I'm
new to. Another possible solution would be to revert the above commit.

-- 
You are receiving this mail because:
You are on the CC list for the bug.