Date: Fri, 31 Oct 2003 07:24:45 +0800 From: David Xu <davidxu@freebsd.org> To: John Baldwin <jhb@freebsd.org> Cc: cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/i386/i386 trap.c src/sys/kern kern_sig.c Message-ID: <3FA19DBD.2050407@freebsd.org> In-Reply-To: <XFMail.20031030114721.jhb@FreeBSD.org> References: <XFMail.20031030114721.jhb@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin wrote: >On 30-Oct-2003 David Xu wrote: > > >>davidxu 2003/10/29 18:55:43 PST >> >> FreeBSD src repository >> >> Modified files: >> sys/i386/i386 trap.c >> sys/kern kern_sig.c >> Log: >> Try to fetch thread mailbox address in page fault trap, so when thread >> blocks in page fault hanlder, and upcall thread can be scheduled. It is >> useful if process is doing lots of mmap based I/O. >> >> > >Does this need to be ported to other arch's? > > > Yes, I think if other arch's want same performance as ia32. Peter has already ported it to amd64. But because I don't know other arch's page fault behavior, for example, is it safe to access a faulting page again in page fault trap if the page is the one includes kse mailbox ? is it possible the arch write half of word in a page fault if the faulting address crosses page boundary ? I need to avoid fetching a not fully written mailbox address, current I do not check the address range on ia32, I think I needn't to do so. David Xu
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3FA19DBD.2050407>