This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [RFA:] Remove extraneous spacing from directives in config/elfos.h


  In message <200009170930.LAA09327@ignucius.axis.se>you write:
  > > Date: Sat, 16 Sep 2000 13:40:03 -0700
  > > From: Nick Clifton <nickc@redhat.com>
  > 
  > > : ! #define CONST_SECTION_ASM_OP	"\t.section\t.rodata"
  > > :   
  > > : --- 254,260 ----
  > > :   
  > > : ! #define CONST_SECTION_ASM_OP	".section\t.rodata"
  > > 
  > > These tabs are all deliberately inserted.  Some assemblers, including
  > > some GAS versions insist that only labels and comments can start in
  > > the first column, and that all other directives, including
  > > instructions and pseudo ops, should be indented.
  > 
  > The change would be from
  > 		.section ....
  > to
  > 	.section ....
  > 
  > *not* to entirely remove the tab.  The ..._ASM_OP users all add
  > a tab AFAICT.  Well except for dbxout.c, whose ..._ASM_OP:s
  > don't contain tabs!
  > 
  > Still doubtful?  If so, tell me and I'll revert my patch locally
  > and run the c++ testsuite and cut and paste those messages and
  > send them.  I didn't just change it on a whim.  Right now, I
  > think the change was simply misunderstood as removing all
  > spacing, putting the dot in the first column.  (Note to self:
  > don't use the word extraneous.)
I'd like to see us decide on a consistent format for this stuff.  ie,
who should include the tab?  The ASM_OP definition or the users of the
ASM_OP.

My gut is telling me the former -- the users should just be outputting
the ASM_OPs as-is with no extra fluff.  BUt I'm willing to go the other
way if someone can make a good case for it.

jeff



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]