Thread (53 messages) 53 messages, 9 authors, 2012-02-29

Re: [PATCH v11 08/12] signal, x86: add SIGSYS info and make it synchronous.

From: Oleg Nesterov <oleg@redhat.com>
Date: 2012-02-27 18:16:27
Also in: lkml, netdev

On 02/27, Roland McGrath wrote:
On Mon, Feb 27, 2012 at 9:22 AM, Oleg Nesterov [off-list ref] wrote:
quoted
SYNCHRONOUS_MASK just tells dequeue_signal() "pick them first".
This is needed to make sure that the handler for, say SIGSEGV,
can use ucontext->ip as a faulting addr.
It's desireable to have these signals handled first just so that the thread
state that provoked the signal is not obscured by an unrelated asynchronous
signal having its handler setup done beforehand.
OK, then probably it makes sense to update the changelog, "To ensure that
SIGSYS delivery occurs on return from the triggering system call" looks
confusing imho.

Not that I really understand why "setup_rt_frame() first" really matters
in this case, siginfo should carry all necessary info. IOW, may be "run
this handler first" makes more sense but this change makes the opposite.

OK, I won't argue, just I was confused by the changelog.

Oleg.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help