revlog: automatically read from opened file handles
The revlog reading code commonly opens a new file handle for
reading on demand. There is support for passing a file handle
to revlog.revision(). But it is marked as an internal argument.
When revlogs are written, we write() data as it is available. But
we don't flush() data until all revisions are written.
Putting these two traits together, it is possible for an in-process
revlog reader during active writes to trigger the opening of a new
file handle on a file with unflushed writes. The reader won't have
access to all "available" revlog data (as it hasn't been flushed).
And with the introduction of the previous patch, this can lead to
the revlog raising an error due to a partial read.
I witnessed this behavior when applying changegroup data (via
`hg pull`) before issue6006 was fixed via different means. Having
this and the previous patch in play would have helped cause errors
earlier rather than manifesting as hash verification failures.
While this has been a long-standing issue, I believe the relatively
new delta computation code has tickled it into being more common.
This is because the new delta computation code will compute deltas
in more scenarios. This can lead to revlog reading. While the delta
computation code is probably supposed to reuse file handles, it
appears it isn't doing so in all circumstances.
But the issue runs deeper than that. Theoretically, any code can
access revision data during revlog writes. It appears we were just
getting lucky that it wasn't. (The "add revision callback" passed to
addgroup() provides an avenue to do this.)
If I changed the revlog's behavior to not cache the full revision
text or to clear caches after revision insertion during addgroup(),
I was able to produce crashes 100% of the time when writing changelog
revisions. This is because changelog's add revision callback attempts
to resolve the revision data to access the changed files list. And
without the revision's fulltext being cached, we performed a revlog
read, which required opening a new file handle. This attempted to read
unflushed data, leading to a partial read and a crash.
This commit teaches the revlog to store the file handles used for
writing multiple revisions during addgroup(). It also teaches the
code for resolving a file handle when reading to use these handles,
if available. This ensures that *any* reads (regardless of their
source) use the active writing file handles, if available. These
file handles have access to the unflushed data because they wrote it.
This allows reads to complete without issue.
Differential Revision: https://phab.mercurial-scm.org/D5267
$ . "$TESTDIR/remotefilelog-library.sh"
Set up an extension to make sure remotefilelog clientsetup() runs
unconditionally even if we have never used a local shallow repo.
This mimics behavior when using remotefilelog with chg. clientsetup() can be
triggered due to a shallow repo, and then the code can later interact with
non-shallow repositories.
$ cat > setupremotefilelog.py << EOF
> from mercurial import extensions
> def extsetup(ui):
> remotefilelog = extensions.find('remotefilelog')
> remotefilelog.onetimeclientsetup(ui)
> EOF
Set up the master repository to pull from.
$ hg init master
$ cd master
$ cat >> .hg/hgrc <<EOF
> [remotefilelog]
> server=True
> EOF
$ echo x > x
$ hg commit -qAm x
$ cd ..
$ hg clone ssh://user@dummy/master child -q
We should see the remotefilelog capability here, which advertises that
the server supports our custom getfiles method.
$ cd master
$ echo 'hello' | hg -R . serve --stdio | grep capa | identifyrflcaps
exp-remotefilelog-ssh-getfiles-1
x_rfl_getfile
x_rfl_getflogheads
$ echo 'capabilities' | hg -R . serve --stdio | identifyrflcaps ; echo
exp-remotefilelog-ssh-getfiles-1
x_rfl_getfile
x_rfl_getflogheads
Pull to the child repository. Use our custom setupremotefilelog extension
to ensure that remotefilelog.onetimeclientsetup() gets triggered. (Without
using chg it normally would not be run in this case since the local repository
is not shallow.)
$ echo y > y
$ hg commit -qAm y
$ cd ../child
$ hg pull --config extensions.setuprfl=$TESTTMP/setupremotefilelog.py
pulling from ssh://user@dummy/master
searching for changes
adding changesets
adding manifests
adding file changes
added 1 changesets with 1 changes to 1 files
new changesets d34c38483be9
(run 'hg update' to get a working copy)
$ hg up
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cat y
y
Test that bundle works in a non-remotefilelog repo w/ remotefilelog loaded
$ echo y >> y
$ hg commit -qAm "modify y"
$ hg bundle --base ".^" --rev . mybundle.hg --config extensions.setuprfl=$TESTTMP/setupremotefilelog.py
1 changesets found
$ cd ..