Mercurial > hg
view tests/test-eol-update.t @ 46607:e9901d01d135
revlog: add a mechanism to verify expected file position before appending
If someone uses `hg debuglocks`, or some non-hg process writes to the .hg
directory without respecting the locks, or if the repo's on a networked
filesystem, it's possible for the revlog code to write out corrupted data.
The form of this corruption can vary depending on what data was written and how
that happened. We are in the "networked filesystem" case (though I've had users
also do this to themselves with the "`hg debuglocks`" scenario), and most often
see this with the changelog. What ends up happening is we produce two items
(let's call them rev1 and rev2) in the .i file that have the same linkrev,
baserev, and offset into the .d file, while the data in the .d file is appended
properly. rev2's compressed_size is accurate for rev2, but when we go to
decompress the data in the .d file, we use the offset that's recorded in the
index file, which is the same as rev1, and attempt to decompress
rev2.compressed_size bytes of rev1's data. This usually does not succeed. :)
When using inline data, this also fails, though I haven't investigated why too
closely. This shows up as a "patch decode" error. I believe what's happening
there is that we're basically ignoring the offset field, getting the data
properly, but since baserev != rev, it thinks this is a delta based on rev
(instead of a full text) and can't actually apply it as such.
For now, I'm going to make this an optional component and default it to entirely
off. I may increase the default severity of this in the future, once I've
enabled it for my users and we gain more experience with it. Luckily, most of my
users have a versioned filesystem and can roll back to before the corruption has
been written, it's just a hassle to do so and not everyone knows how (so it's a
support burden). Users on other filesystems will not have that luxury, and this
can cause them to have a corrupted repository that they are unlikely to know how
to resolve, and they'll see this as a data-loss event. Refusing to create the
corruption is a much better user experience.
This mechanism is not perfect. There may be false-negatives (racy writes that
are not detected). There should not be any false-positives (non-racy writes that
are detected as such). This is not a mechanism that makes putting a repo on a
networked filesystem "safe" or "supported", just *less* likely to cause
corruption.
Differential Revision: https://phab.mercurial-scm.org/D9952
author | Kyle Lippincott <spectral@google.com> |
---|---|
date | Wed, 03 Feb 2021 16:33:10 -0800 |
parents | e88549a02f5e |
children | 87faafd6ea83 |
line wrap: on
line source
Test EOL update $ cat >> $HGRCPATH <<EOF > [diff] > git = 1 > EOF $ seteol () { > if [ $1 = "LF" ]; then > EOL='\n' > else > EOL='\r\n' > fi > } $ makerepo () { > echo > echo "# ==== setup repository ====" > echo '% hg init' > hg init repo > cd repo > > cat > .hgeol <<EOF > [patterns] > **.txt = LF > EOF > > printf "first\nsecond\nthird\n" > a.txt > printf "f\r\n" > f > hg commit --addremove -m 'LF commit' > > cat > .hgeol <<EOF > [patterns] > **.txt = CRLF > f = LF > EOF > > printf "first\r\nsecond\r\nthird\r\n" > a.txt > printf "f\n" > f > hg commit -m 'CRLF commit' > > cd .. > } $ dotest () { > seteol $1 > > echo > echo "% hg clone repo repo-$1" > hg clone --noupdate repo repo-$1 > cd repo-$1 > > cat > .hg/hgrc <<EOF > [extensions] > eol = > EOF > > hg update > > echo '% a.txt (before)' > cat a.txt > > printf "first${EOL}third${EOL}" > a.txt > > echo '% a.txt (after)' > cat a.txt > echo '% hg diff' > hg diff > > echo '% hg update 0' > hg update 0 > > echo '% a.txt' > cat a.txt > echo '% hg diff' > hg diff > > > cd .. > rm -r repo-$1 > } $ makerepo # ==== setup repository ==== % hg init adding .hgeol adding a.txt adding f $ dotest LF % hg clone repo repo-LF 3 files updated, 0 files merged, 0 files removed, 0 files unresolved % a.txt (before) first\r (esc) second\r (esc) third\r (esc) % a.txt (after) first third % hg diff diff --git a/a.txt b/a.txt --- a/a.txt +++ b/a.txt @@ -1,3 +1,2 @@ first\r (esc) -second\r (esc) third\r (esc) % hg update 0 merging a.txt 2 files updated, 1 files merged, 0 files removed, 0 files unresolved % a.txt first third % hg diff diff --git a/a.txt b/a.txt --- a/a.txt +++ b/a.txt @@ -1,3 +1,2 @@ first -second third $ dotest CRLF % hg clone repo repo-CRLF 3 files updated, 0 files merged, 0 files removed, 0 files unresolved % a.txt (before) first\r (esc) second\r (esc) third\r (esc) % a.txt (after) first\r (esc) third\r (esc) % hg diff diff --git a/a.txt b/a.txt --- a/a.txt +++ b/a.txt @@ -1,3 +1,2 @@ first\r (esc) -second\r (esc) third\r (esc) % hg update 0 merging a.txt 2 files updated, 1 files merged, 0 files removed, 0 files unresolved % a.txt first third % hg diff diff --git a/a.txt b/a.txt --- a/a.txt +++ b/a.txt @@ -1,3 +1,2 @@ first -second third Test in repo using eol extension, while keeping an eye on how filters are applied: $ cd repo $ hg up -q -c -r null $ cat > .hg/hgrc <<EOF > [extensions] > eol = > EOF Update to revision 0 which has no .hgeol, shouldn't use any filters, and obviously should leave things as tidy as they were before the clean update. $ hg up -c -r 0 -v --debug resolving manifests branchmerge: False, force: False, partial: False ancestor: 000000000000, local: 000000000000+, remote: 15cbdf8ca3db calling hook preupdate.eol: hgext.eol.preupdate .hgeol: remote created -> g getting .hgeol filtering .hgeol through isbinary a.txt: remote created -> g getting a.txt filtering a.txt through tolf f: remote created -> g getting f 3 files updated, 0 files merged, 0 files removed, 0 files unresolved $ hg st $ hg branch b marked working directory as branch b (branches are permanent and global, did you want a bookmark?) $ hg ci -m b Merge changes that apply a filter to f: $ hg merge 1 3 files updated, 0 files merged, 0 files removed, 0 files unresolved (branch merge, don't forget to commit) $ hg st M .hgeol M a.txt M f $ hg diff diff --git a/.hgeol b/.hgeol --- a/.hgeol +++ b/.hgeol @@ -1,2 +1,3 @@ [patterns] -**.txt = LF +**.txt = CRLF +f = LF diff --git a/a.txt b/a.txt --- a/a.txt +++ b/a.txt @@ -1,3 +1,3 @@ -first -second -third +first\r (esc) +second\r (esc) +third\r (esc) diff --git a/f b/f --- a/f +++ b/f @@ -1,1 +1,1 @@ -f\r (esc) +f Abort the merge with up -C to revision 0. Note that files are filtered correctly for revision 0: f is not filtered, a.txt is filtered with tolf, and everything is left tidy. $ touch .hgeol * # ensure consistent dirtyness checks ignoring dirstate $ hg up -C -r 0 -v --debug eol: detected change in .hgeol resolving manifests branchmerge: False, force: True, partial: False ancestor: 1db78bdd3bd6+, local: 1db78bdd3bd6+, remote: 15cbdf8ca3db calling hook preupdate.eol: hgext.eol.preupdate .hgeol: remote is newer -> g getting .hgeol filtering .hgeol through isbinary a.txt: remote is newer -> g getting a.txt filtering a.txt through tolf f: remote is newer -> g getting f 3 files updated, 0 files merged, 0 files removed, 0 files unresolved $ touch .hgeol * $ hg st --debug eol: detected change in .hgeol filtering .hgeol through isbinary filtering a.txt through tolf skip updating dirstate: identity mismatch (?) $ hg diff Things were clean, and updating again will not change anything: $ touch .hgeol * $ hg up -C -r 0 -v --debug eol: detected change in .hgeol filtering .hgeol through isbinary filtering a.txt through tolf resolving manifests branchmerge: False, force: True, partial: False ancestor: 15cbdf8ca3db+, local: 15cbdf8ca3db+, remote: 15cbdf8ca3db calling hook preupdate.eol: hgext.eol.preupdate starting 4 threads for background file closing (?) 0 files updated, 0 files merged, 0 files removed, 0 files unresolved $ touch .hgeol * $ hg st --debug eol: detected change in .hgeol filtering .hgeol through isbinary filtering a.txt through tolf $ cd .. $ rm -r repo