diff options
author | James Zern <jzern@google.com> | 2020-02-14 21:00:21 -0800 |
---|---|---|
committer | James Zern <jzern@google.com> | 2020-02-14 21:03:56 -0800 |
commit | 0f3fe088fab5998c37910f3cc539a48fa1a72c14 (patch) | |
tree | 17d6c343ddb5549402e7c84e8ac885d5f93eb6f6 /test/test-data.mk | |
parent | 9cfcac1cb3b2e765c24706a9e966859d588bc4ef (diff) | |
download | libvpx-0f3fe088fab5998c37910f3cc539a48fa1a72c14.tar libvpx-0f3fe088fab5998c37910f3cc539a48fa1a72c14.tar.gz libvpx-0f3fe088fab5998c37910f3cc539a48fa1a72c14.tar.bz2 libvpx-0f3fe088fab5998c37910f3cc539a48fa1a72c14.zip |
vp8_decode: add missing vpx_clear_system_state
this avoids leaving the floating point unit in an inconsistent state on
error and breaking subsequent tests on x86
the test clip invalid-bug-148271109.ivf would also result in a sanitizer
error prior to:
vp8,GetSigned: silence unsigned int overflow warning
BUG=b/148271109
Change-Id: Ia254f3892ac1eeec51db5e9d42ea071545db0cd8
Diffstat (limited to 'test/test-data.mk')
-rw-r--r-- | test/test-data.mk | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/test/test-data.mk b/test/test-data.mk index 905f0138e..81f035d83 100644 --- a/test/test-data.mk +++ b/test/test-data.mk @@ -737,6 +737,8 @@ endif # CONFIG_VP9_HIGHBITDEPTH # Invalid files for testing libvpx error checking. LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-bug-1443.ivf LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-bug-1443.ivf.res +LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-bug-148271109.ivf +LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-bug-148271109.ivf.res LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-token-partition.ivf LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-token-partition.ivf.res LIBVPX_TEST_DATA-$(CONFIG_VP8_DECODER) += invalid-vp80-00-comprehensive-018.ivf.2kf_0x6.ivf |