Is there any good reason for using DW_FORM_data1 for the DW_AT_language
attribute of DW_TAG_compile_unit? This prevents using language codes in
the DW_LANG_lo_user=0x8000 to DW_LANG_hi_user=0xffff range.
I think the obvious change to use DW_FORM_data2 in line 1897 of
lib/CodeGen/AsmPrinter/DwarfDebug.cpp should fix it. Or, if there are
binary compatibility issues, the abbrev form could be chosen
conditionally. Are there?
Is there any good reason for using DW_FORM_data1 for the DW_AT_language
attribute of DW_TAG_compile_unit? This prevents using language codes in
the DW_LANG_lo_user=0x8000 to DW_LANG_hi_user=0xffff range.
I think the obvious change to use DW_FORM_data2 in line 1897 of
lib/CodeGen/AsmPrinter/DwarfDebug.cpp should fix it. Or, if there are
binary compatibility issues, the abbrev form could be chosen
conditionally. Are there?
this bit be when I tried to use a language ID in the user range
> I think the obvious change to use DW_FORM_data2 in line 1897 of
> lib/CodeGen/AsmPrinter/DwarfDebug.cpp should fix it. Or, if there are
> binary compatibility issues, the abbrev form could be chosen
> conditionally. Are there?
I do not see any binary compatibility issue. In any case, if there are
any I'd learn them about soon! Checked in change to use DW_FORM_data2. r126339.