[Bug 213936] stable/11 -r307797 on BPi-M3 (cortex-a7): toolchain produces bad stack code in lang/gcc6's xgcc's cc1 and so cc1 gets SIGSYS during lang/gcc6 bootstrap(?)

bugzilla-noreply at freebsd.org bugzilla-noreply at freebsd.org
Tue Nov 1 19:19:35 UTC 2016


https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=213936

Mark Millard <markmi at dsl-only.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |Unable to Reproduce
             Status|New                         |Closed

--- Comment #3 from Mark Millard <markmi at dsl-only.net> ---
I tried rebuilding lang/gcc6 and it completed instead of stopping like it did
before.

So apparently the USB SSD involved glitched during the original build attempt
--or some other such non-repeating issue happened.

The original bad (xgcc's) cc1 code still shows problems with truss handling of
SIGSYS and odd (huge) syscall numbers reported by ktrace (and internally to
truss as seen via gdb). But there are separate bugzilla reports for those
issues.

-- 
You are receiving this mail because:
You are the assignee for the bug.


More information about the freebsd-arm mailing list