This test is problematic for the bots, since it can behave differently some
configurations, including crashing.
See the following (probably bogus) stack trace:
[ RUN ] BufferDataTest/1.HugeSetDataShouldNotCrash
Backtrace:
glWaitSync [0x6DC50B8E+174]
glWaitSync [0x6DCBDC32+446802]
glWaitSync [0x6DCBDE75+447381]
glWaitSync [0x6DCBDD3C+447068]
glWaitSync [0x6DC65590+84656]
gl::BufferData [0x6DC43ECB+171]
(No symbol) [0x0011F3EC]
(No symbol) [0x002370C8]
(No symbol) [0x002371F9]
(No symbol) [0x002377C9]
(No symbol) [0x00237516]
(No symbol) [0x00208DED]
(No symbol) [0x000F37FB]
(No symbol) [0x002063D7]
(No symbol) [0x00206676]
(No symbol) [0x000F3E79]
(No symbol) [0x0029F16E]
BaseThreadInitThunk [0x76A1338A+18]
RtlInitializeExceptionChain [0x7765BF32+99]
RtlInitializeExceptionChain [0x7765BF05+54]
[52/431] BufferDataTest/1.HugeSetDataShouldNotCrash (CRASHED)
We should redesign it to use mocks, or some reproducible method.
Original issue reported on code.google.com by jmad...@chromium.org on 20 Jan 2015 at 9:46
Original issue reported on code.google.com by
jmad...@chromium.org
on 20 Jan 2015 at 9:46