Date: Wed, 6 Mar 2019 17:25:46 +0000 From: Edward Napierala <trasz@freebsd.org> To: Konstantin Belousov <kib@freebsd.org> Cc: Alexander Leidinger <Alexander@leidinger.net>, Kirk McKusick <mckusick@mckusick.com>, FreeBSD FS <freebsd-fs@freebsd.org> Subject: Re: 'td' vs sys/fs/nfsserver/ Message-ID: <CAFLM3-pbRNq1Hz%2BSXWsEr22PR7Kdh9sno2mwDUjN0HkG7G2k_g@mail.gmail.com> In-Reply-To: <20190306145031.GC2492@kib.kiev.ua> References: <201903052008.x25K8d4Z011653@chez.mckusick.com> <169532d2770.27fa.fa4b1493b064008fe79f0f905b8e5741@Leidinger.net> <20190306145031.GC2492@kib.kiev.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 6 Mar 2019 at 14:50, Konstantin Belousov <kib@freebsd.org> wrote: > > On Wed, Mar 06, 2019 at 02:25:16PM +0100, Alexander Leidinger via freebsd-fs wrote: [..] > > To make it explicit what my intend is with those questions: I suggest to > > look at real numbers and quality, instead of expectations and feelings. > Good luck measuring frustration. We can try to estimate it, though. Actually removing the 'td' argument throughout the kernel would not only introduce code churn (which is a valid concern of course), but it would create one huge merge conflict (or, if spread over time, a series of merge conflicts) for literally everyone. That's a lot of frustration. On the other hand, I'd expect cleaning up the top of the stack to make it easy to stop passing the td in newly written code - by making it very clear, and perhaps assertable, that for a given KPI it's always equal to curthread - to be more acceptable. It would also make it possible, in future, to remove the td in selected places, where it eg causes a stack spill.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAFLM3-pbRNq1Hz%2BSXWsEr22PR7Kdh9sno2mwDUjN0HkG7G2k_g>