https://sourceware.org/bugzilla/show_bug.cgi?id=28791
--- Comment #6 from Nick Clifton <nickc at redhat dot com> --- Created attachment 14024 --> https://sourceware.org/bugzilla/attachment.cgi?id=14024&action=edit Proposed Patch (In reply to Pete Moore from comment #5) > Since the commit above is the one that introduced the bug - should it be > reverted? Hi Peter, I do not think that will be necessary. The commit adds a reasonable test to the assembler. Instead how about this patch which tweaks the test so that it does not complain about constants that are generated through bit manipulation operators (eg ~ and -) ? It fixes the problem as reported in this PR, although it does also mean that the assembler would not warn about overlarge values that are also subject to manipulation, eg ~0x12345678. Cheers Nick -- You are receiving this mail because: You are on the CC list for the bug.
