[Bug 103183] [CI] igt at perf@short-reads -fail - Failed assertion: ret == -1

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=103183

--- Comment #1 from mwa <matthew.auld at intel.com> ---
8bytes is exactly the size of the report header, so probably just a case of
being unlucky with the read() giving us a REPORT_LOST, which just so happens to
fit into the non-protected region so we don't get an EFAULT.

First thought would just be:

diff --git a/tests/perf.c b/tests/perf.c
index ad62319e..fe8e32bf 100644
--- a/tests/perf.c
+++ b/tests/perf.c
@@ -3121,7 +3121,13 @@ test_short_reads(void)
        /* A read that can't return a single record because it would result
         * in a fault on buffer overrun should result in an EFAULT error...
         */
-       ret = read(stream_fd, pages + page_size - 16, page_size);
+       do {
+               header = (void *)(pages + page_size - 16);
+               ret = read(stream_fd,
+                          header,
+                          page_size);
+       } while (header->type == DRM_I915_PERF_RECORD_OA_REPORT_LOST);
+
        igt_assert_eq(ret, -1);
        igt_assert_eq(errno, EFAULT);

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are on the CC list for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20171018/43464fb4/attachment.html>