[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <d87e3b4dfd4624d182d3d23992eacb7b9ffeff90.camel@gmail.com>
Date: Fri, 27 Sep 2024 16:22:01 -0700
From: Eduard Zingerman <eddyz87@...il.com>
To: David Howells <dhowells@...hat.com>
Cc: Manu Bretelle <chantr4@...il.com>, asmadeus@...ewreck.org,
ceph-devel@...r.kernel.org, christian@...uner.io, ericvh@...nel.org,
hsiangkao@...ux.alibaba.com, idryomov@...il.com, jlayton@...nel.org,
linux-afs@...ts.infradead.org, linux-cifs@...r.kernel.org,
linux-erofs@...ts.ozlabs.org, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-mm@...ck.org,
linux-nfs@...r.kernel.org, marc.dionne@...istor.com,
netdev@...r.kernel.org, netfs@...ts.linux.dev, pc@...guebit.com,
smfrench@...il.com, sprasad@...rosoft.com, tom@...pey.com,
v9fs@...ts.linux.dev, willy@...radead.org
Subject: Re: [PATCH v2 19/25] netfs: Speed up buffered reading
On Fri, 2024-09-27 at 22:11 +0100, David Howells wrote:
[...]
> If you look here:
>
> https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs.git/log/?h=netfs-fixes
>
> you can see some patches I've added. If you can try this branch or cherry
> pick:
>
> netfs: Fix write oops in generic/346 (9p) and generic/074 (cifs)
> netfs: Advance iterator correctly rather than jumping it
> netfs: Use a folio_queue allocation and free functions
> netfs: Add a tracepoint to log the lifespan of folio_queue structs
I used your branch netfs-fixes, namely at the following commit:
8e18fe180b0a ("netfs: Abstract out a rolling folio buffer implementation")
> And then turn on the following "netfs" tracepoints:
>
> read,sreq,rreq,failure,write,write_iter,folio,folioq,progress,donate
>
System can't boot, so I used the following kernel command line:
... trace_event=:netfs_read,:netfs_sreq,:netfs_rreq,:netfs_failure,:netfs_write,:netfs_write_iter,:netfs_folio,:netfs_folioq,:netfs_progress,:netfs_donate
No warnings like "Failed to enable trace event ...", so I assume it worked
as expected.
A fresh dmesg is here:
https://gist.github.com/eddyz87/e8f4780d833675a7e58854596394a70f
Don't see any tracepoint output there, so something is probably missing.
> > Alternatively I can pack this thing in a dockerfile, so that you would
> > be able to reproduce locally (but that would have to wait till my evening).
>
> I don't have Docker set up, so I'm not sure how easy that would be for me to
> use.
What's your preferred setup for the repro?
Powered by blists - more mailing lists