Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upSIGILL with --multiprocess #7755
Comments
oesiman
commented
Sep 27, 2015
|
Curious! I can't reproduce this on OS X in a debug build. What platform are you using? Do you see a backtrace from the first panic ( |
|
Alternatively, if you could run it under your favourite debugger, set a breakpoint on |
|
On Sun, Sep 27, 2015 at 06:31:05AM -0700, Josh Matthews wrote:
|
|
I dug a little into this, and I don't see an obvious bounds check in either serde_macros, the |
|
I wasn't able to reproduce this on the master branch. @oesiman: could you see if it's still reproducible? |
|
On Tue, Sep 29, 2015 at 05:34:45PM -0700, Erick Tryzelaar wrote:
I can reproduce it (but i'm sure that i must be wrong...). about my system: I tried two master again on: [master] Auto merge of #7719 - Manishearth:ghfix, r=Ms2ger I name #7719 "old" and #7790 "new": running from shell gives SIGILL: RUST_BACKTRACE=1 ./servo --multiprocess http://heise.de running under gdb (overkill): http://txt.hello-penguin.com/servo/gdb-old.txt.xz I also put the servo-binaries on the net if somebody ciao - |
|
Since it's been 5 years, this should probably be closed. |
|
We run multiprocess tests regularly without panics, so closing makes sense until it can be reproduced. |