Hi Al,
I spotted a possible anomaly with ARC sigaltstack() handling and was wondering if
you could confirm if it was wrong (I don't have a test case to prove one way or
the other).
We have a micro-optimization which elides calling restore_altstack() in sigreturn
path, based on a "magic" value, set at the time of sigaction (SA_SIGINFO). So
essentially we only restore ss_{sp,size,flags} if SA_SIGINFO was set.
Do you think this needs to be done unconditionally like other ports do ?
Thx,
-Vineet