> -----Original Message----- > From: gcc-patches-ow...@gcc.gnu.org [mailto:gcc-patches- > ow...@gcc.gnu.org] On Behalf Of Matthew Fortune > Sent: Thursday, November 17, 2016 8:47 AM > To: Toma Tabacu <toma.tab...@imgtec.com>; Andrew Bennett > <andrew.benn...@imgtec.com>; Moore, Catherine > <catherine_mo...@mentor.com>; 'gcc-patches@gcc.gnu.org' <gcc- > patc...@gcc.gnu.org> > Subject: RE: [PATCH] MIPS: If a test in the MIPS testsuite requires > standard library support check the sysroot supports the required test > options. > > Toma Tabacu <toma.tab...@imgtec.com> writes: > > Hi, > > > > The patch below is a rebased version of Andrew's patch plus a few > more changes > > to test options. > > > > I have tested Andrew's patch by passing -msoft-float to the testsuite > without > > having a soft-float library available, and saw that the inline-memcpy- > {1,2,3,4,5}.c > > and memcpy-1.c tests were also failing to find standard library > headers. > > In the version below, I have added (REQUIRES_STDLIB) to them as > well. > > > > However, I believe that the memcpy-1.c test was removed from the > first version > > of Andrew's patch in response to Matthew's comments, but I don't > think it > > should be. > > > > Tested with mips-img-linux-gnu and mips-mti-linux-gnu. > > This looks OK to me but I then again I helped with the design for this. > > I'd like to give Catherine a chance to take a look though as the feature > is unusual. > > One comment below. > > > diff --git a/gcc/testsuite/gcc.target/mips/mips.exp > > b/gcc/testsuite/gcc.target/mips/mips.exp > > index e22d782..ccd4ecb 100644 > > --- a/gcc/testsuite/gcc.target/mips/mips.exp > > +++ b/gcc/testsuite/gcc.target/mips/mips.exp > > @@ -1420,6 +1426,22 @@ proc mips-dg-options { args } { > > } > > } > > > > + # If the test is marked as requiring standard libraries check > > + # that the sysroot has support for the current set of test options. > > + if { [mips_have_test_option_p options "REQUIRES_STDLIB"] } { > > + mips_push_test_options saved_options $extra_tool_flags > > + set output [mips_preprocess "" { > > + #include <stdlib.h> > > + } 1] > > + mips_pop_test_options saved_options > > + > > + # If the preprocessing of the stdlib.h file produced errors mark > > + # the test as unsupported. > > + if { ![string equal $output ""] } { > > + set do_what [lreplace $do_what 1 1 "N"] > > We should describe what we expect the format of do_what to be at > the time > of writing in case it ever changes. i.e. a comment to say what the > second > character means and therefore make it clear that setting it to 'n' makes > the test unsupported. >
This patch looks okay to me after updating the comment as Matthew suggested.
