This is the mail archive of the gdb-patches@sourceware.cygnus.com mailing list for the GDB project.


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

Re: Patch to fix thread support for Solaris x86


   Date: Mon, 28 Feb 2000 11:54:15 +1100
   From: Andrew Cagney <ac131313@cygnus.com>

   Jim Kingdon wrote:
   > 
   > > This patch fixes thread support for Solaris x86 platforms.
   > 
   > Looks like a straightforward bug fix.
   > 
   > Maintainer, as with the other Solaris x86 patch, currently unknown.  I
   > like Kevin Buettner's suggestion of appointing a generic x86
   > maintainer, if we can find a victim.

   At present I'm checking them in after a thumbs up from JimB and/or
   MarkK.

Peter's patch has a small problem (for which I am probably to blame,
since it appears I suggested it): it conditionalizes the code on
HAVE_I387_REGS, which according to the following comment in
i386-tdep.c isn't right:

   /* This register file is parameterized by two macros:
      HAVE_I387_REGS --- register file should include i387 registers
      HAVE_SSE_REGS  --- register file should include SSE registers
      If HAVE_SSE_REGS is #defined, then HAVE_I387_REGS must also be #defined.

      However, GDB code should not test those macros with #ifdef, since
      that makes code which is annoying to multi-arch.  Instead, GDB code
      should check the values of NUM_GREGS, NUM_FREGS, and NUM_SSE_REGS,
      which will eventually get mapped onto architecture vector entries.

      It's okay to use the macros in tm-*.h files, though, since those
      files will get completely replaced when we multi-arch anyway.  */

Can you fix that Peter?

I'm also not sure about the usefullness of the testsuite change.  The
absence of the i387 registers on a i386 target probably means that the
maintainer has some work to do :-).  A reminder in the testsuite
wouldn't hurt too much I guess.

   As for x86 maintainer, the following positions are notionally available:

	   o	x86 target architecture

	   o	Solaris/x86 native/host

		   I've had one suggestion here.

	   o	*BSD/x86 native/host

   The x86 target architecture person would preferably be someone fairly
   native netural and more experienced with target/embedded and multi-arch
   stuff (1).  Mark, how to you feel with respect to those criteria?

I have absolutely no experience with embedded targets.  I have been
hacking on GDB for Hurd/i386 and Linux/i386, and I do have an old i486
with FreeBSD installed.  I think I have a basic understanding of the
most important multi-arch issues, but I would probably need some
guidance if we ever go that route for x86.  

Mark

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