From: David Howells <dhowells@redhat.com>
To: Leon Romanovsky <leon@kernel.org>
Cc: dhowells@redhat.com, Eduard Zingerman <eddyz87@gmail.com>,
Christian Brauner <brauner@kernel.org>,
Manu Bretelle <chantr4@gmail.com>,
asmadeus@codewreck.org, ceph-devel@vger.kernel.org,
christian@brauner.io, ericvh@kernel.org,
hsiangkao@linux.alibaba.com, idryomov@gmail.com,
jlayton@kernel.org, linux-afs@lists.infradead.org,
linux-cifs@vger.kernel.org, linux-erofs@lists.ozlabs.org,
linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
linux-mm@kvack.org, linux-nfs@vger.kernel.org,
marc.dionne@auristor.com, netdev@vger.kernel.org,
netfs@lists.linux.dev, pc@manguebit.com, smfrench@gmail.com,
sprasad@microsoft.com, tom@talpey.com, v9fs@lists.linux.dev,
willy@infradead.org
Subject: Re: [PATCH v2 19/25] netfs: Speed up buffered reading
Date: Mon, 30 Sep 2024 13:44:30 +0100 [thread overview]
Message-ID: <2968940.1727700270@warthog.procyon.org.uk> (raw)
In-Reply-To: <20240925103118.GE967758@unreal>
Okay, let's try something a little more drastic. See if we can at least get
it booting to the point we can read the tracelog. If you can apply the
attached patch? It won't release any folio_queue struct or put the refs on
any pages, so it will quickly run out of memory - but if you have sufficient
menory, it might be enough to boot.
David
---
9p: [DEBUGGING] Don't release pages or folioq structs
diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c
index af46a598f4d7..702286484176 100644
--- a/fs/netfs/buffered_read.c
+++ b/fs/netfs/buffered_read.c
@@ -84,8 +84,8 @@ static size_t netfs_load_buffer_from_ra(struct netfs_io_request *rreq,
folioq->orders[i] = order;
size += PAGE_SIZE << order;
- if (!folio_batch_add(put_batch, folio))
- folio_batch_release(put_batch);
+ //if (!folio_batch_add(put_batch, folio))
+ // folio_batch_release(put_batch);
}
for (int i = nr; i < folioq_nr_slots(folioq); i++)
diff --git a/fs/netfs/misc.c b/fs/netfs/misc.c
index 63280791de3b..cec55b7eb5bc 100644
--- a/fs/netfs/misc.c
+++ b/fs/netfs/misc.c
@@ -88,7 +88,7 @@ struct folio_queue *netfs_delete_buffer_head(struct netfs_io_request *wreq)
if (next)
next->prev = NULL;
netfs_stat_d(&netfs_n_folioq);
- kfree(head);
+ //kfree(head);
wreq->buffer = next;
return next;
}
@@ -108,11 +108,11 @@ void netfs_clear_buffer(struct netfs_io_request *rreq)
continue;
if (folioq_is_marked(p, slot)) {
trace_netfs_folio(folio, netfs_folio_trace_put);
- folio_put(folio);
+ //folio_put(folio);
}
}
netfs_stat_d(&netfs_n_folioq);
- kfree(p);
+ //kfree(p);
}
}
next prev parent reply other threads:[~2024-09-30 12:44 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-08-14 20:38 [PATCH v2 00/25] netfs: Read/write improvements David Howells
2024-08-14 20:38 ` [PATCH v2 01/25] netfs, ceph: Partially revert "netfs: Replace PG_fscache by setting folio->private and marking dirty" David Howells
2024-08-14 20:38 ` [PATCH v2 02/25] cachefiles: Fix non-taking of sb_writers around set/removexattr David Howells
2024-08-14 20:38 ` [PATCH v2 03/25] netfs: Adjust labels in /proc/fs/netfs/stats David Howells
2024-08-14 20:38 ` [PATCH v2 04/25] netfs: Record contention stats for writeback lock David Howells
2024-08-14 20:38 ` [PATCH v2 05/25] netfs: Reduce number of conditional branches in netfs_perform_write() David Howells
2024-08-14 20:38 ` [PATCH v2 06/25] netfs, cifs: Move CIFS_INO_MODIFIED_ATTR to netfs_inode David Howells
2024-08-14 20:38 ` [PATCH v2 07/25] netfs: Move max_len/max_nr_segs from netfs_io_subrequest to netfs_io_stream David Howells
2024-08-14 20:38 ` [PATCH v2 08/25] netfs: Reserve netfs_sreq_source 0 as unset/unknown David Howells
2024-08-14 20:38 ` [PATCH v2 09/25] netfs: Remove NETFS_COPY_TO_CACHE David Howells
2024-08-14 20:38 ` [PATCH v2 10/25] netfs: Set the request work function upon allocation David Howells
2024-08-14 20:38 ` [PATCH v2 11/25] netfs: Use bh-disabling spinlocks for rreq->lock David Howells
2024-08-14 20:38 ` [PATCH v2 12/25] mm: Define struct folio_queue and ITER_FOLIOQ to handle a sequence of folios David Howells
2024-08-14 20:38 ` [PATCH v2 13/25] iov_iter: Provide copy_folio_from_iter() David Howells
2024-08-14 20:38 ` [PATCH v2 14/25] cifs: Provide the capability to extract from ITER_FOLIOQ to RDMA SGEs David Howells
2024-08-14 20:38 ` [PATCH v2 15/25] netfs: Use new folio_queue data type and iterator instead of xarray iter David Howells
2024-09-24 9:48 ` Leon Romanovsky
2024-08-14 20:38 ` [PATCH v2 16/25] netfs: Provide an iterator-reset function David Howells
2024-08-14 20:38 ` [PATCH v2 17/25] netfs: Simplify the writeback code David Howells
2024-08-14 20:38 ` [PATCH v2 18/25] afs: Make read subreqs async David Howells
2024-08-14 20:38 ` [PATCH v2 19/25] netfs: Speed up buffered reading David Howells
2024-08-16 11:12 ` Simon Horman
2024-09-23 18:34 ` Manu Bretelle
2024-09-23 18:43 ` Eduard Zingerman
2024-09-23 21:56 ` Eduard Zingerman
2024-09-23 22:33 ` David Howells
2024-09-23 19:38 ` David Howells
2024-09-23 20:20 ` Manu Bretelle
2024-09-24 23:20 ` David Howells
2024-09-25 0:01 ` Eduard Zingerman
2024-09-25 10:31 ` Leon Romanovsky
2024-09-29 9:12 ` David Howells
2024-09-29 9:37 ` Eduard Zingerman
2024-09-29 18:55 ` Leon Romanovsky
2024-09-30 12:44 ` David Howells [this message]
2024-09-30 12:51 ` David Howells
2024-09-30 16:46 ` Eduard Zingerman
2024-09-30 18:35 ` David Howells
2024-09-30 19:00 ` Omar Sandoval
2024-09-27 20:50 ` David Howells
2024-09-27 20:55 ` Eduard Zingerman
2024-09-27 21:11 ` David Howells
2024-09-27 23:22 ` Eduard Zingerman
2024-08-14 20:38 ` [PATCH v2 20/25] netfs: Remove fs/netfs/io.c David Howells
2024-08-14 20:38 ` [PATCH v2 21/25] cachefiles, netfs: Fix write to partial block at EOF David Howells
2024-08-14 20:38 ` [PATCH v2 22/25] netfs: Cancel dirty folios that have no storage destination David Howells
2024-08-14 20:38 ` [PATCH v2 23/25] cifs: Use iterate_and_advance*() routines directly for hashing David Howells
2024-08-14 20:38 ` [PATCH v2 24/25] cifs: Switch crypto buffer to use a folio_queue rather than an xarray David Howells
2024-08-14 20:38 ` [PATCH v2 25/25] cifs: Don't support ITER_XARRAY David Howells
2024-08-15 13:07 ` [PATCH v2 00/25] netfs: Read/write improvements Christian Brauner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2968940.1727700270@warthog.procyon.org.uk \
--to=dhowells@redhat.com \
--cc=asmadeus@codewreck.org \
--cc=brauner@kernel.org \
--cc=ceph-devel@vger.kernel.org \
--cc=chantr4@gmail.com \
--cc=christian@brauner.io \
--cc=eddyz87@gmail.com \
--cc=ericvh@kernel.org \
--cc=hsiangkao@linux.alibaba.com \
--cc=idryomov@gmail.com \
--cc=jlayton@kernel.org \
--cc=leon@kernel.org \
--cc=linux-afs@lists.infradead.org \
--cc=linux-cifs@vger.kernel.org \
--cc=linux-erofs@lists.ozlabs.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-nfs@vger.kernel.org \
--cc=marc.dionne@auristor.com \
--cc=netdev@vger.kernel.org \
--cc=netfs@lists.linux.dev \
--cc=pc@manguebit.com \
--cc=smfrench@gmail.com \
--cc=sprasad@microsoft.com \
--cc=tom@talpey.com \
--cc=v9fs@lists.linux.dev \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).