This is the mail archive of the gcc-bugs@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]
Other format: [Raw text]

other/6480: C testsuite '--tool_opts' has incorrect precedence



>Number:         6480
>Category:       other
>Synopsis:       C testsuite '--tool_opts' has incorrect precedence
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Apr 26 11:06:01 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     Kelley Cook <kelleycook@comcast.net>
>Release:        gcc version 3.1 20020423 (prerelease)
>Organization:
>Environment:
i686-pc-cygwin
>Description:
When invesigating the difference between testsuite runs with and without RUNFLAGS="--tool_opts='-march=pentium3 -mfpmath=sse'", it turned out that all the failures were either due to poorly constructed testcases or from a bug in C testing framework.

The latter occured because gcc.exp, unlike g77.exp and g++.exp, appends tool_opts *after* the Testcase-specific options.  Some tests may require those particular --tool_opt options to be overriden.
>How-To-Repeat:
make -k RUNTESTFLAGS="--tool_opts '-march=pentium3'" check gcc

Specifically, look at the failure of gcc.misc-tests/i386-pf-none-1.c
>Fix:
See Attachment for one-line patch.  It simply puts the --tool_opts overrides before any test specific options.

Testsuite behaves identically without "--tool_opts" specified.

I don't use Objective C, but the same patch could be applied to objc.exp.
>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/octet-stream; name="gcc.exp.diff"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="gcc.exp.diff"
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Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]