On Wed 21 Nov 2007 09:02, Bernd Schmidt pondered: > [EMAIL PROTECTED] wrote: > > > + /* Ensure that bad return addresses don't end up in an infinite > > + * loop, due to speculative loads/reads. This needs to be done after > > + * the signal has been sent. > > + */ > > + if (trapnr == VEC_CPLB_I_M && sig != SIGTRAP) > > + fp->pc = SAFE_USER_INSTRUCTION; > > + > > Do we have a testcase for the infinite loop? I'm slightly worried that > even after this checkin it may be causing confusion in user space.
I found the problem that caused the infinite loop - and fixed it in the kernel - so you are correct - and I removed that code. If there is anything else that causes a crash when it shouldn't, we can add it to the dist user/blkfin-test/crash_test/traps_test.c and then debug it from there. -Robin _______________________________________________ Linux-kernel-commits mailing list [email protected] http://blackfin.uclinux.org/mailman/listinfo/linux-kernel-commits
