Before this patch, upgrade will process everything, re-clone all revlogs, write
requirements file again even there is no change to be made.
This patch makes it exit earlier.
Details
Details
- Reviewers
mharbison72 - Group Reviewers
hg-reviewers - Commits
- rHGd3113c4cf52c: upgrade: don't perform anything if nothing to do
Diff Detail
Diff Detail
- Repository
- rHG Mercurial
- Branch
- default
- Lint
No Linters Available - Unit
No Unit Test Coverage
Event Timeline
Comment Actions
I get test failures with this:
$ ./run-tests.py --local test-upgrade-repo.t running 1 tests using 1 parallel processes --- /mnt/c/Users/Matt/hg/tests/test-upgrade-repo.t +++ /mnt/c/Users/Matt/hg/tests/test-upgrade-repo.t.err @@ -1317,17 +1317,6 @@ downgrade $ hg debugupgraderepo --run --no-backup --quiet - upgrade will perform the following actions: - - requirements - preserved: dotencode, fncache, generaldelta, revlogv1, sparserevlog, store - removed: revlog-compression-zstd - - processed revlogs: - - all-filelogs - - changelog - - manifest - $ hg debugformat -v format-variant repo config default fncache: yes yes yes @@ -1339,12 +1328,13 @@ persistent-nodemap: no no no copies-sdc: no no no plain-cl-delta: yes yes yes - compression: zlib zlib zlib + compression: zstd zlib zlib compression-level: default default default $ cat .hg/requires dotencode fncache generaldelta + revlog-compression-zstd revlogv1 sparserevlog store @@ -1356,17 +1346,6 @@ > revlog-compression=zstd > EOF $ hg debugupgraderepo --run --no-backup --quiet - upgrade will perform the following actions: - - requirements - preserved: dotencode, fncache, generaldelta, revlogv1, sparserevlog, store - added: revlog-compression-zstd - - processed revlogs: - - all-filelogs - - changelog - - manifest - $ hg debugformat -v format-variant repo config default fncache: yes yes yes ERROR: test-upgrade-repo.t output changed ! Failed test-upgrade-repo.t: output changed # Ran 1 tests, 0 skipped, 1 failed. python hash seed: 3820465322
I have no idea why the compression difference here. Something else to be aware of (not related to this commit) is that test-upgrade-repo.t is extremely flaky on Windows. If I run it by itself, it usually fails in a rename and then has cascading errors. If I run it with test-check-* in the list, it may or may not fail the same way. For example:
--- c:/Users/Matt/hg/tests/test-upgrade-repo.t +++ c:/Users/Matt/hg/tests/test-upgrade-repo.t.err @@ -944,75 +944,69 @@ starting in-place swap of repository data replaced files will be backed up at $TESTTMP/upgradegd/.hg/upgradebackup.* (glob) replacing store... - store replacement complete; repository was inconsistent for *s (glob) - finalizing requirements file and making repository readable again - removing old repository content $TESTTMP/upgradegd/.hg/upgradebackup.* (glob) - removing temporary repository $TESTTMP/upgradegd/.hg/upgrade.* (glob) + removing temporary repository $TESTTMP\upgradegd\.hg\upgrade.qzp1er + Traceback (most recent call last): + File "c:\Users\Matt\hg\mercurial\scmutil.py", line 155, in callcatch + return func() + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 455, in _runcatchfunc + return _dispatch(req) + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 1252, in _dispatch + lui, repo, cmd, fullargs, ui, options, d, cmdpats, cmdoptions + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 906, in runcommand + ret = _runcommand(ui, options, cmd, d) + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 1263, in _runcommand + return cmdfunc() + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 1249, in <lambda> + d = lambda: util.checksignature(func)(ui, *args, **strcmdopt) + File "c:\Users\Matt\hg\mercurial\util.py", line 1867, in check + return func(*args, **kwargs) + File "c:\Users\Matt\hg\mercurial\debugcommands.py", line 3936, in debugupgraderepo + ui, repo, run=run, optimize=set(optimize), backup=backup, **opts + File "c:\Users\Matt\hg\mercurial\upgrade.py", line 206, in upgraderepo + ui, repo, dstrepo, upgrade_op + File "c:\Users\Matt\hg\mercurial\upgrade_utils\engine.py", line 512, in upgrade + _replacestores(srcrepo, dstrepo, backupvfs, upgrade_op) + File "c:\Users\Matt\hg\mercurial\upgrade_utils\engine.py", line 427, in _replacestores + util.rename(upgradedrepo.spath, currentrepo.spath) + File "c:\Users\Matt\hg\mercurial\windows.py", line 603, in rename + os.rename(src, dst) + WindowsError: [Error 5] Access is denied + abort: Access is denied + [255] $ hg verify - checking changesets - checking manifests - crosschecking files in changesets and manifests - checking files - checked 3 changesets with 3 changes to 3 files + abort: repository requires features unknown to this Mercurial: upgradeinprogress + (see https://mercurial-scm.org/wiki/MissingRequirement for more information) + [255] Check you can't skip revlog clone during important format upgrade ...
Comment Actions
Yeah, one of parent patch which needs change fixes this issue. Once D9693 is pushed, the failure will go away.
I have no idea why the compression difference here. Something else to be aware of (not related to this commit) is that test-upgrade-repo.t is extremely flaky on Windows. If I run it by itself, it usually fails in a rename and then has cascading errors. If I run it with test-check-* in the list, it may or may not fail the same way. For example:
--- c:/Users/Matt/hg/tests/test-upgrade-repo.t +++ c:/Users/Matt/hg/tests/test-upgrade-repo.t.err @@ -944,75 +944,69 @@ starting in-place swap of repository data replaced files will be backed up at $TESTTMP/upgradegd/.hg/upgradebackup.* (glob) replacing store... - store replacement complete; repository was inconsistent for *s (glob) - finalizing requirements file and making repository readable again - removing old repository content $TESTTMP/upgradegd/.hg/upgradebackup.* (glob) - removing temporary repository $TESTTMP/upgradegd/.hg/upgrade.* (glob) + removing temporary repository $TESTTMP\upgradegd\.hg\upgrade.qzp1er + Traceback (most recent call last): + File "c:\Users\Matt\hg\mercurial\scmutil.py", line 155, in callcatch + return func() + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 455, in _runcatchfunc + return _dispatch(req) + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 1252, in _dispatch + lui, repo, cmd, fullargs, ui, options, d, cmdpats, cmdoptions + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 906, in runcommand + ret = _runcommand(ui, options, cmd, d) + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 1263, in _runcommand + return cmdfunc() + File "c:\Users\Matt\hg\mercurial\dispatch.py", line 1249, in <lambda> + d = lambda: util.checksignature(func)(ui, *args, **strcmdopt) + File "c:\Users\Matt\hg\mercurial\util.py", line 1867, in check + return func(*args, **kwargs) + File "c:\Users\Matt\hg\mercurial\debugcommands.py", line 3936, in debugupgraderepo + ui, repo, run=run, optimize=set(optimize), backup=backup, **opts + File "c:\Users\Matt\hg\mercurial\upgrade.py", line 206, in upgraderepo + ui, repo, dstrepo, upgrade_op + File "c:\Users\Matt\hg\mercurial\upgrade_utils\engine.py", line 512, in upgrade + _replacestores(srcrepo, dstrepo, backupvfs, upgrade_op) + File "c:\Users\Matt\hg\mercurial\upgrade_utils\engine.py", line 427, in _replacestores + util.rename(upgradedrepo.spath, currentrepo.spath) + File "c:\Users\Matt\hg\mercurial\windows.py", line 603, in rename + os.rename(src, dst) + WindowsError: [Error 5] Access is denied + abort: Access is denied + [255]
Oh, this is error-ing out in final stage of upgrade where we swap data. Since I have a windows machine now, I will try to check what's going on.
$ hg verify
- checking changesets
- checking manifests
- crosschecking files in changesets and manifests
- checking files
- checked 3 changesets with 3 changes to 3 files
+ abort: repository requires features unknown to this Mercurial: upgradeinprogress
+ (see https://mercurial-scm.org/wiki/MissingRequirement for more information)
+ [255]
Check you can't skip revlog clone during important format upgrade
...
Comment Actions
I mostly support this change. However, a supposed no-op upgrade may still result in differences! For example, if we change the algorithm for computing how revlog deltas are computed without actually changing the storage semantics, performing a no-op upgrade would migrate the store to use the new code.
For this reason, we probably want to have a --force option or something to override the default no-op behavior.