Date: Tue, 26 Jun 2001 11:12:51 -0700 (PDT) From: Matt Dillon <dillon@earth.backplane.com> To: "Jonathan Lemon <jlemon@flugsvamp.com> Alfred Perlstein" <bright@sneakerz.org> Cc: "Brian F. Feldman" <green@FreeBSD.ORG>, Mike Silbersack <silby@FreeBSD.ORG>, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG, jlemon@FreeBSD.ORG, bmilekic@FreeBSD.ORG Subject: Re: cvs commit: src/sys/netinet tcp_input.c tcp_output.c tcp_subr.c tcp_timer.c tcp_usrreq.c tcp_var.h Message-ID: <200106261812.f5QICpr26493@earth.backplane.com> References: <200106242141.f5OLfc176777@green.bikeshed.org> <200106261647.f5QGlRr24342@earth.backplane.com> <20010626121752.L64836@sneakerz.org>
next in thread | previous in thread | raw e-mail | index | archive | help
:Agreed, wrt idle zeroing of data, the problem that I see is that
:it seems that idleproc may not aquire mutexes, at least that's the
:impression I get from the SMP team (jhb and crew). I feel that
:idleproc should at least be able to trylock if we're going to make
:any use of it at all.
:
:Is this planned or possible?
:
:-Alfred
idle_proc() can only safely obtain spin mutexes, and do non-blocking mutex
operations. The trylock() equivalent should be ok but I haven't looked
at it closely. idle_proc() cannot ever block though I suppose it could
be interrupted and switched out preemptively.
The only reason I #if 0'd out the vm_page_zero_idle() code in idle_proc()
was because I didn't have time to clean up its locking or move
vm_page_zero_idle() from the MD section to the MI section (where it
belongs). At this stage in the game it should be fairly easy for you
guys to get that working again, and maybe even start working on
a pre-zeroing feature for malloc/M_ZERO :-)
-Matt
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200106261812.f5QICpr26493>