Re: RFA: Test ability to access i386 SSE registers

[Jim Blandy <jimb@redhat.com>]

Jim Blandy <jimb@redhat.com> writes:

> Daniel Jacobowitz <drow@false.org> writes:
> 
> > On Wed, May 12, 2004 at 12:25:40AM -0500, Jim Blandy wrote:
> > > 
> > > It's late; there is surely some Kool-Aid around here I have forgotten
> > > to drink.
> > 
> > There better be!
> > 
> > > + if ![runto_main] then {
> > > +     gdb_suppress_tests
> > > + }
> > > + 
> > > + gdb_test "break [gdb_get_line_number "first breakpoint here"]" \
> > > +          "Breakpoint .* at .*i386-sse.c.*" \
> > > +          "set breakpoint in main"
> > > + gdb_continue_to_breakpoint "continue to first breakpoint in main"
> > 
> > So, all that trouble to make the SSE testcase not use movaps if it
> > isn't available... and you never check in the test script?
> 
> Oh yeah.  That was it.  [vacant stare]
> 
> > If you're going to be thorough in the test case, be more thorough.  See
> > gcc.dg/i386-cpuid.h in GCC.
> 
> Okay.  How about:
> 
> 2004-05-12  Jim Blandy  <jimb@redhat.com>
> 
> 	* gdb.arch/i386-sse.exp, gdb.arch/i386-sse.c: New tests.
> 	* gdb.arch/i386-cpuid.h: New helper file.

I've committed this test as revised.