This is the mail archive of the
mailing list for the GDB project.
Re: [RFA] Improve performance with lots of shared libraries
- From: Tom Tromey <tromey at redhat dot com>
- To: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- Cc: Pedro Alves <pedro at codesourcery dot com>, gdb-patches at sourceware dot org, Daniel Jacobowitz <drow at false dot org>
- Date: Tue, 04 Oct 2011 13:46:17 -0600
- Subject: Re: [RFA] Improve performance with lots of shared libraries
- References: <20110909193239.GA23130@host1.jankratochvil.net> <email@example.com> <20110912160436.GA16320@host1.jankratochvil.net> <20110914092037.GA15919@redhat.com>
>>>>> "Gary" == Gary Benson <firstname.lastname@example.org> writes:
Pedro> Only one, when Tom's work is done, make breakpoints have a "final"
Pedro> property (meaning, "stop adding locations for this breakpoint")
Pedro> for user breakpoints, and automatically mark some internal
Pedro> breakpoints we know can't have more than one location with that
Pedro> flag. Breakpoints on addresses ("b *FOO") would have the "final"
Pedro> property implicitly as well. If all the breakpoints in the table
Pedro> are "final", then we don't need to track shared library loads.
Jan> BTW I agree with this design.
Gary> Me too.
Gary> I will leave this patch on ice until Tom's work is done, or at least
Gary> until Tom is back and can comment.
I haven't read the patch yet, just this subthread.
I also agree with this approach; in fact I think it is what we all
agreed to in the long thread determining the new semantics.