>-----Original Message----- >From: Daniel Vetter [mailto:daniel.vetter@xxxxxxxx] On Behalf Of Daniel Vetter >Sent: Thursday, May 28, 2015 10:05 AM >To: Morton, Derek J >Cc: intel-gfx@xxxxxxxxxxxxxxxxxxxxx; Wood, Thomas >Subject: Re: [PATCH i-g-t v3] lib/tests/igt_segfault.c Add unit test to test segfault handling > >On Wed, May 27, 2015 at 04:54:24PM +0100, Derek Morton wrote: >> Unit test to check a segfaulting subtest is handled correctly. >> >> v2: Added script to check subtest results >> v3: Removed script. Updated test to use fork to monitor return status. >> >> Signed-off-by: Derek Morton <derek.j.morton@xxxxxxxxx> >> --- >> lib/tests/Makefile.sources | 1 + >> lib/tests/igt_segfault.c | 139 +++++++++++++++++++++++++++++++++++++++++++++ >> 2 files changed, 140 insertions(+) >> create mode 100644 lib/tests/igt_segfault.c >> >> diff --git a/lib/tests/Makefile.sources b/lib/tests/Makefile.sources >> index 10e0617..5fa0b31 100644 >> --- a/lib/tests/Makefile.sources >> +++ b/lib/tests/Makefile.sources >> @@ -8,6 +8,7 @@ check_PROGRAMS = \ >> igt_simple_test_subtests \ >> igt_timeout \ >> igt_invalid_subtest_name \ >> + igt_segfault \ >> $(NULL) >> >> check_SCRIPTS = \ >> diff --git a/lib/tests/igt_segfault.c b/lib/tests/igt_segfault.c new >> file mode 100644 index 0000000..b420b1a >> --- /dev/null >> +++ b/lib/tests/igt_segfault.c >> @@ -0,0 +1,139 @@ >> +/* >> + * Copyright (c) 2015 Intel Corporation >> + * >> + * Permission is hereby granted, free of charge, to any person >> +obtaining a >> + * copy of this software and associated documentation files (the >> +"Software"), >> + * to deal in the Software without restriction, including without >> +limitation >> + * the rights to use, copy, modify, merge, publish, distribute, >> +sublicense, >> + * and/or sell copies of the Software, and to permit persons to whom >> +the >> + * Software is furnished to do so, subject to the following conditions: >> + * >> + * The above copyright notice and this permission notice (including >> +the next >> + * paragraph) shall be included in all copies or substantial portions >> +of the >> + * Software. >> + * >> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, >> +EXPRESS OR >> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF >> +MERCHANTABILITY, >> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT >> +SHALL >> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES >> +OR OTHER >> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, >> +ARISING >> + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR >> +OTHER DEALINGS >> + * IN THE SOFTWARE. >> + * >> + * Authors: >> + * Derek Morton <derek.j.morton@xxxxxxxxx> >> + * >> + */ >> + >> +/* >> + * Testcase: Test the framework catches a segfault and returns an error. >> + * >> + * 1. Test a crashing simple test is reported. >> + * 2. Test a crashing subtest is reported. >> + * 3. Test a crashing subtest following a passing subtest is reported. >> + * 4. Test a crashing subtest preceeding a passing subtest is reported. >> + */ >> + >> +#include <stdlib.h> >> +#include <sys/wait.h> >> +#include <sys/types.h> >> +#include <assert.h> >> +#include <errno.h> >> + >> +#include "drmtest.h" >> +#include "igt_core.h" >> + >> +/* >> + * We need to hide assert from the cocci igt test refactor spatch. >> + * >> + * IMPORTANT: Test infrastructure tests are the only valid places >> +where using >> + * assert is allowed. >> + */ >> +#define internal_assert assert >> + >> +bool simple; >> +bool runa; >> +bool runc; >> +char test[] = "test"; >> +char *argv_run[] = { test }; >> + >> +static int do_fork(void) >> +{ >> + int pid, status; >> + int argc; >> + void (*crashme)(void) = NULL; >> + >> + switch (pid = fork()) { >> + case -1: >> + internal_assert(0); >> + case 0: >> + if (simple) { >> + argc = 1; >> + igt_simple_init(argc, argv_run); >> + crashme(); >> + >> + igt_exit(); >> + } else { >> + >> + argc = 1; >> + igt_subtest_init(argc, argv_run); >> + >> + if(runa) >> + igt_subtest("A") >> + ; >> + >> + igt_subtest("B") >> + crashme(); >> + >> + if(runc) >> + igt_subtest("C") >> + ; >> + >> + igt_exit(); >> + } >> + default: >> + while (waitpid(pid, &status, 0) == -1 && >> + errno == EINTR) >> + ; >> + >> + if(WIFSIGNALED(status)) >> + return WTERMSIG(status) + 128; >> + >> + return WEXITSTATUS(status); >> + } >> +} >> + >> +int main(int argc, char **argv) >> +{ >> + /* Test Crash in simple test is reported */ >> + simple = true; >> + runa=false; >> + runc=false; >> + igt_info("Simple test.\n"); >> + fflush(stdout); >> + internal_assert(do_fork() == SIGSEGV + 128); >> + >> + /* Test crash in a single subtest is reported */ >> + simple = false; >> + igt_info("Single subtest.\n"); >> + fflush(stdout); >> + internal_assert(do_fork() == SIGSEGV + 128); > >This one here fails for me when running make check. I didn't look into details. Also something seemed off with the content-encoding of your mail, at least git apply-mbox complained. I didn't debug but applied it manually. > >Also please add igt_segfault to the relevant .gitignore. >-Daniel > I submitted a separate patch "[PATCH i-g-t] lib/igt_core.c: Flag the test as failing after a segfault" as segfaults were not being flagged as a failure. If that patch has not yet been merged I would expect this test to fail at this point. (I was asked to write this UT when I submitted that patch.) Git send-email asked about encoding when I submitted the patch. I assumed it was the (c) character in the copyright header. I just selected whatever it recommended. (UTF8 I think). Previous patches have not been a problem. I will update the .gitignore and resubmit this patch. //Derek >> + >> + /* Test crash in a subtest following a pass is reported */ >> + simple = false; >> + runa=true; >> + igt_info("Passing then crashing subtest.\n"); >> + fflush(stdout); >> + internal_assert(do_fork() == SIGSEGV + 128); >> + >> + /* Test crash in a subtest preceeding a pass is reported */ >> + simple = false; >> + runa=false; >> + runc=true; >> + igt_info("Crashing then passing subtest.\n"); >> + fflush(stdout); >> + internal_assert(do_fork() == SIGSEGV + 128); >> + >> + return 0; >> +} >> + >> -- >> 1.9.1 >> > > >-- >Daniel Vetter >Software Engineer, Intel Corporation >http://blog.ffwll.ch > _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
