Control: retitle -1 gtest-port_test fails on m68k On Friday, December 9, 2016 12:40:59 PM CST David Kalnischkies wrote: > On Wed, Dec 07, 2016 at 11:28:09PM -0600, Steve M. Robbins wrote:
> Given that this is proven to be false by now I think we can all move on > more or less pretending it never happened – so from my PoV feel free to > close the bugreport or take it as m68k port-bug. I think John is > actually involved in m68k porting, so he might be able to lend a hand! Agree: I'll leave it open as an m68k port bug. What is your opinion as to bug severity? > More than I can at least as I can only provide a pad on the back and > a "that sounds strange indeed". Although, exceptions are a tricky beast > and "costly", so I see why a compiler would like to optimize them out if > given the chance… but that is not really helping. Does upstream know > about it? Also, Debian isn't the only distro with ports – we might > 'only' be the ones with the most – so in theory others should see such > issues, too. To clarify: the exception test failure happens on amd64, so it should be visible in essentially any other distribution. Dmitri did report it upstream [1] but there has been no response. My googling for "gtest_catch_exceptions_test failure" only turned up the Debian discussions. This is what makes me suspect something about the Debian tool chain. -Steve [1] https://github.com/google/googletest/issues/845
signature.asc
Description: This is a digitally signed message part.
