mm/filemap: drop check for truncated page after I/O
authorMatthew Wilcox (Oracle) <willy@infradead.org>
Fri, 30 Apr 2021 05:55:29 +0000 (22:55 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 30 Apr 2021 18:20:37 +0000 (11:20 -0700)
commit79e3094c53c56d0d4da23f578de271e7602ba5ed
treef8851511eef12c9a0a1bf80c03d687a2047faa40
parentd31fa86a27b3ecdc32bf19326c4d3bba854542e2
mm/filemap: drop check for truncated page after I/O

If the I/O completed successfully, the page will remain Uptodate, even
if it is subsequently truncated.  If the I/O completed with an error,
this check would cause us to retry the I/O if the page were truncated
before we woke up.  There is no need to retry the I/O; the I/O to fill
the page failed, so we can legitimately just return -EIO.

This code was originally added by commit 56f0d5fe6851 ("[PATCH]
readpage-vs-invalidate fix") in 2005 (this commit ID is from the
linux-fullhistory tree; it is also commit ba1f08f14b52 in tglx-history).

At the time, truncate_complete_page() called ClearPageUptodate(), and so
this was fixing a real bug.  In 2008, commit 84209e02de48 ("mm: dont clear
PG_uptodate on truncate/invalidate") removed the call to
ClearPageUptodate, and this check has been unnecessary ever since.

It doesn't do any real harm, but there's no need to keep it.

Link: https://lkml.kernel.org/r/20210303222547.1056428-1-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: William Kucharski <william.kucharski@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/filemap.c