Mercurial > hg
view tests/test-debugbuilddag.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 | 88e7105b5cd9 |
children |
line wrap: on
line source
plain $ hg init $ hg debugbuilddag '+2:f +3:p2 @temp <f+4 @default /p2 +2' \ > --config extensions.progress= --config progress.assume-tty=1 \ > --config progress.delay=0 --config progress.refresh=0 \ > --config progress.format=topic,bar,number \ > --config progress.width=60 \r (no-eol) (esc) building [ ] 0/12\r (no-eol) (esc) building [ ] 0/12\r (no-eol) (esc) building [==> ] 1/12\r (no-eol) (esc) building [==> ] 1/12\r (no-eol) (esc) building [======> ] 2/12\r (no-eol) (esc) building [=========> ] 3/12\r (no-eol) (esc) building [=============> ] 4/12\r (no-eol) (esc) building [=============> ] 4/12\r (no-eol) (esc) building [=============> ] 4/12\r (no-eol) (esc) building [================> ] 5/12\r (no-eol) (esc) building [====================> ] 6/12\r (no-eol) (esc) building [=======================> ] 7/12\r (no-eol) (esc) building [===========================> ] 8/12\r (no-eol) (esc) building [===========================> ] 8/12\r (no-eol) (esc) building [==============================> ] 9/12\r (no-eol) (esc) building [==================================> ] 10/12\r (no-eol) (esc) building [=====================================> ] 11/12\r (no-eol) (esc) \r (no-eol) (esc) tags $ cat .hg/localtags 66f7d451a68b85ed82ff5fcc254daf50c74144bd f bebd167eb94d257ace0e814aeb98e6972ed2970d p2 dag $ hg debugdag -t -b +2:f +3:p2 @temp*f+3 @default*/p2+2:tip tip $ hg id 000000000000 glog $ hg log -G --template '{rev}: {desc} [{branches}] @ {date}\n' o 11: r11 [] @ 11.00 | o 10: r10 [] @ 10.00 | o 9: r9 [] @ 9.00 |\ | o 8: r8 [temp] @ 8.00 | | | o 7: r7 [temp] @ 7.00 | | | o 6: r6 [temp] @ 6.00 | | | o 5: r5 [temp] @ 5.00 | | o | 4: r4 [] @ 4.00 | | o | 3: r3 [] @ 3.00 | | o | 2: r2 [] @ 2.00 |/ o 1: r1 [] @ 1.00 | o 0: r0 [] @ 0.00 overwritten files, starting on a non-default branch $ rm -r .hg $ hg init $ hg debugbuilddag '@start.@default.:f +3:p2 @temp <f+4 @default /p2 +2' -q -o tags $ cat .hg/localtags f778700ebd50fcf282b23a4446bd155da6453eb6 f bbccf169769006e2490efd2a02f11c3d38d462bd p2 dag $ hg debugdag -t -b @start+1 @default+1:f +3:p2 @temp*f+3 @default*/p2+2:tip tip $ hg id 000000000000 glog $ hg log -G --template '{rev}: {desc} [{branches}] @ {date}\n' o 11: r11 [] @ 11.00 | o 10: r10 [] @ 10.00 | o 9: r9 [] @ 9.00 |\ | o 8: r8 [temp] @ 8.00 | | | o 7: r7 [temp] @ 7.00 | | | o 6: r6 [temp] @ 6.00 | | | o 5: r5 [temp] @ 5.00 | | o | 4: r4 [] @ 4.00 | | o | 3: r3 [] @ 3.00 | | o | 2: r2 [] @ 2.00 |/ o 1: r1 [] @ 1.00 | o 0: r0 [start] @ 0.00 glog of $ hg log -G --template '{rev}: {desc} [{branches}]\n' of o 11: r11 [] | o 10: r10 [] | o 9: r9 [] |\ | o 8: r8 [temp] | | | o 7: r7 [temp] | | | o 6: r6 [temp] | | | o 5: r5 [temp] | | o | 4: r4 [] | | o | 3: r3 [] | | o | 2: r2 [] |/ o 1: r1 [] | o 0: r0 [start] tags $ hg tags -v tip 11:9ffe238a67a2 p2 4:bbccf1697690 local f 1:f778700ebd50 local cat of $ hg cat of --rev tip r11 new and mergeable files $ rm -r .hg $ hg init $ hg debugbuilddag '+2:f +3:p2 @temp <f+4 @default /p2 +2' -q -mn dag $ hg debugdag -t -b +2:f +3:p2 @temp*f+3 @default*/p2+2:tip tip $ hg id 000000000000 glog $ hg log -G --template '{rev}: {desc} [{branches}] @ {date}\n' o 11: r11 [] @ 11.00 | o 10: r10 [] @ 10.00 | o 9: r9 [] @ 9.00 |\ | o 8: r8 [temp] @ 8.00 | | | o 7: r7 [temp] @ 7.00 | | | o 6: r6 [temp] @ 6.00 | | | o 5: r5 [temp] @ 5.00 | | o | 4: r4 [] @ 4.00 | | o | 3: r3 [] @ 3.00 | | o | 2: r2 [] @ 2.00 |/ o 1: r1 [] @ 1.00 | o 0: r0 [] @ 0.00 glog mf $ hg log -G --template '{rev}: {desc} [{branches}]\n' mf o 11: r11 [] | o 10: r10 [] | o 9: r9 [] |\ | o 8: r8 [temp] | | | o 7: r7 [temp] | | | o 6: r6 [temp] | | | o 5: r5 [temp] | | o | 4: r4 [] | | o | 3: r3 [] | | o | 2: r2 [] |/ o 1: r1 [] | o 0: r0 [] man r4 $ hg manifest -r4 mf nf0 nf1 nf2 nf3 nf4 cat r4 mf $ hg cat -r4 mf 0 r0 1 2 r1 3 4 r2 5 6 r3 7 8 r4 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 man r8 $ hg manifest -r8 mf nf0 nf1 nf5 nf6 nf7 nf8 cat r8 mf $ hg cat -r8 mf 0 r0 1 2 r1 3 4 5 6 7 8 9 10 r5 11 12 r6 13 14 r7 15 16 r8 17 18 19 20 21 22 23 man $ hg manifest --rev tip mf nf0 nf1 nf10 nf11 nf2 nf3 nf4 nf5 nf6 nf7 nf8 nf9 cat mf $ hg cat mf --rev tip 0 r0 1 2 r1 3 4 r2 5 6 r3 7 8 r4 9 10 r5 11 12 r6 13 14 r7 15 16 r8 17 18 r9 19 20 r10 21 22 r11 23