This is the mail archive of the
mailing list for the GDB project.
- From: palves at sourceware dot org
- To: gdb-cvs at sourceware dot org
- Date: 9 Sep 2011 19:29:52 -0000
- Subject: src/gdb ChangeLog
Module name: src
Changes by: email@example.com 2011-09-09 19:29:52
gdb : ChangeLog
Fixup last entry.
2011-09-09 Pedro Alves <firstname.lastname@example.org>
* linux-nat.h (enum resume_kind): New.
(struct lwp_info) <last_resume_kind>: New field.
* linux-nat.c (linux_child_follow_fork): Set last_resume_kind to
resume_stop on the new lwp.
(add_lwp): Set last_resume_kind as resume_continue by default.
(lin_lwp_attach_lwp): Set last_resume_kind as resume_stop.
(resume_lwp): New, factored out from resume_callback. Also check
for pending status in lp->waitstatus.
(resume_clear_callback): Set last_resume_kind as resume_stop.
(resume_set_callback): Set last_resume_kind as resume_continue.
(linux_nat_resume, linux_handle_extended_wait): Set
(running_callback): Also check lp->waitstatus for pending events.
(select_singlestep_lwp_callback): Check that lp->last_resume_kind
(stop_and_resume_callback): Don't re-resume if the core wanted the
lwp stopped. Use resume_lwp instead of resume_callback. Avoid
using an invalidated pointer.
(linux_nat_filter_event): Don't discard SIGSTOPs as delayed
SIGSTOPs if the core wanted the LWP to stop.
(linux_nat_wait_1) Don't consume a pending SIGSTOP if the core
wanted the lwp to stop. If the core wanted the lwp to stop, and
the lwp stopped with a SIGSTOP, report a TARGET_SIGNAL_0 instead
(linux_nat_stop_lwp): Don't synchronously wait for the lwp to stop
here. Instead, signal the lwp, and set the last_resume_kind to