| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Change-Id: I1f6928de56895db956cd02d78a546629dfe4cd1c
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Issue:
When LFS operation is initiated through the SSH LFS client receives auth
token and uses it to perform requested operation e.g.:
POST /gerrit/test-org/test-no-block/info/lfs/objects/batch
Authorization: Ssh: ...
Content-Type: application/vnd.git-lfs+json; charset=utf-8
{
"operation":"upload","objects"...
}
ContainerAuthFilter searches for existing user but none of the
containers can perform successful LFS auth (as it is deeply buried in
the plugin internals) therefore typically it is configured to let it go
through to eventually fail in the filter with:
403 Forbidden
Solution:
Modify ContainerAuthFilter so that it returns 'true' when Content-Type
indicates LFS request is against LFS and Authorization header value
starts with "Ssh: " string (similarly to ProjectBasicAuthFilter when
it doesn't start with "Basic ").
Rationale:
ContainerAuthFilter is installed for requests that either go through
/a/* (authorised path) or to LFS (note that LFS over HTTP sends
auth token even when request is not `/a/` prefixed - hence user
can be obtained from request with the first call without sending
401 back and re-sending request with `/a/` prefix.
In terms of LFS over SSH it is again request against LFS but in this
case it has `Ssh: ` based auth token that is not recognized by
filter and results in 403.
This change is safe as it introduces exception only for LFS requests
that rely on internal LFS auth anyway.
Change-Id: Ia886dc284c8ded9c21a5b9f57628f228c1e691f0
Signed-off-by: Jacek Centkowski <jcentkowski@collab.net>
|
|
|
|
|
|
|
|
|
|
|
| |
According to the release notes [1] there are no REST client specific
fixes. This upgrade is only to keep up-to-date with the latest version.
Also update the test container to use this version for the 6.6 tests.
[1] https://www.elastic.co/guide/en/elasticsearch/reference/current/release-notes-6.6.1.html
Change-Id: Ie17d0402e0f7b6c02da6875cf2f765ba2f2ad7ec
|
|
|
|
|
|
|
|
|
| |
According to the release notes [1], there are no changes in this new
version. It is likely just to align with other Elastic stack update(s).
[1] https://www.elastic.co/guide/en/elasticsearch/reference/5.6/release-notes-5.6.15.html
Change-Id: I345e9121a45a6c478a3b3707409924ad232d0239
|
|\
| |
| |
| |
| |
| | |
* changes:
AccountIT: Extend assertKeyEquals to check that full key is included
Fix GPG public key export
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Assert that the full public key is included in the output, i.e. it is
not truncated before the "-----END PGP PUBLIC KEY BLOCK-----" footer.
The new assertion fails without the fix previously added in Id9569d969.
Bug: Issue 10488
Change-Id: Idbd00c0d138be0a6616b782fd747ca156d79fce4
|
| |
| |
| |
| |
| |
| |
| |
| | |
The armored output stream needs to be closed to finish the armoring process.
Failure to do so leads to incomplete GPG public keys transferred to the UI/REST API.
Bug: Issue 10488
Change-Id: Id9569d969d362ea93a8b8259fd05abe141a05dfd
|
| |
| |
| |
| |
| |
| |
| | |
This updates it to mention that both wip and private have "patchSet"
now.
Change-Id: If84a717f37934cb545f5dfcfe99b74e56bd47606
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Update plugins/hooks from branch 'stable-2.15'
to 491c5ddf0e3892ffe684cc825a406ef56744a458
- Merge "Include 'submit' hook in listing of hooks run synchronously" into stable-2.15
- Include 'submit' hook in listing of hooks run synchronously
Appears to be omitted on config.md - see the hooks.md where this is listed
as hook with an ability to prevent a submit, before this has happened.
Change-Id: I446f05c1f2372c92b3bb3dcf2d81a64958c2c3b6
|
| |
| |
| |
| |
| |
| |
| |
| | |
* Update plugins/hooks from branch 'stable-2.15'
to 657f7c53e2fc331f7c85ea0f609b314f5b1f95ee
- Fix typo on hooks listing page
Change-Id: Ic3802d99cc1d3f2057830d6d7d53e8f5fa963397
|
|\ \ |
|
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Previously, in order to remove an access rule, an empty object could not
be the value for the rule. This change adds a check for 'action' and
so that it does not thow a null pointer exception.
Example:
Throws an error:
{
add: {},
remove: {'refs/heads/*: {
permissions: {
label-code-review: {
rules: {xxx: {}}
}
}
}
}
Does not throw an error:
{
add: {},
remove: {'refs/heads/*: {
permissions: {
label-code-review: {
rules: {xxx: null}
}
}
}
}
Other entities (ex: permissions) require an object for removal
Example:
Throws an error:
{
add: {},
remove: {'refs/heads/*: {
permissions: {
label-code-review: null
}
}
}
Does not throw an error:
{
add: {},
remove: {'refs/heads/*: {
permissions: {
label-code-review: {}
}
}
}
Bug: Issue 10106
Change-Id: Id6c4e21bcf5964421ed976f4531d2cf294082e37
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | | |
* stable-2.14:
setup_gjf.sh: Add support for version 1.7
Change-Id: Ibf258f496cbfd7c6573350ce0d52747f5e406641
|
| | |
| | |
| | |
| | | |
Change-Id: I0eb1533a85b4002af17cc355c7d20074861f6b31
|
|\ \ \ |
|
| | |/
| |/|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
As listed in the "Breaking changes" section of the release notes [1],
index creation requests are rejected when the include_type_name
parameter is not set to "true". This is a new parameter introduced
in 7.0.0, so we set it conditionally only for that version.
[1] https://www.elastic.co/guide/en/elasticsearch/reference/7.0/release-notes-7.0.0-beta1.html
Change-Id: I71ac74567eba11cf4150619a2cb95dce4386ac0f
|
|\ \ \
| | |/
| |/|
| | |
| | |
| | |
| | | |
* stable-2.14:
gerrit.sh: Improve error message when JRE cannot be found
Change-Id: I6f70bdafe7dc364463cad07a9f900ff126b7d306
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If a JRE cannot be found at the location specified by JAVA_HOME or
container.javaHome, the server startup fails and an error message is
displayed.
Update the error message so that the reader realizes that while the
JAVA_HOME and/or container.javaHome might be *set* the value in them
might not be *valid*. An example of when this might be the case is when
the java installation was upgraded and the location changed from:
/Library/Java/JavaVirtualMachines/jdk1.8.0_91.jdk/Contents/Home/jre
to:
/Library/Java/JavaVirtualMachines/jdk1.8.0_192.jdk/Contents/Home/jre
and while the JAVA_HOME environment variable got update to the new
location, container.javaHome did not and still pointed to the old (and
now invalid) location.
Change-Id: I24e0dc19781920e8901ef8743c847606ebd640ce
|
|\| |
| |/
|/|
| |
| |
| |
| | |
* stable-2.14:
PrologEnvironment#setPredicate: Reduce log level to DEBUG
Change-Id: I2ec9564a7be7f27bd65e9e04e1d96bdfe3667471
|
| |
| |
| |
| |
| |
| | |
Logging at INFO causes too much log spam.
Change-Id: I254bd8865bc6d880e8e1b6b491a8ce325021999a
|
|\|
| |
| |
| |
| |
| |
| | |
* stable-2.14:
PrologEnvironment: Add logging of reduction limits
Change-Id: I9e9a9ce1d82de958ba06dabd3f29714deb52945d
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In the constructor of the Args singleton, log the configured values
for reductionLimit and compileLimit.
In setPredicate, log the reduction limit that is actually used.
Change-Id: I300735ba4a7982de89337a5c179149023894acb6
|
|\|
| |
| |
| |
| |
| |
| | |
* stable-2.14:
Allow dash in name of junit_tests rule
Change-Id: I9a3c5d06fd443c9a90a18b87e685599f21383b38
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
So far, if the name of the rule contained a dash (hyphen), as in
high-availability plugin for example, the test execution failed with
error message:
bazel-out/local-fastbuild/bin/high-availability_testsTestSuite.java:6:
error: '{' expected
public class high-availability_testsTestSuite {}
^
This is due to the fact that dashes are not valid characters in the name
of a Java class.
Workaround this by replacing any dash with an underscore.
Change-Id: I80fd4c5bcff0d2af6f337636bd942f30d6a521c6
|
|\|
| |
| |
| |
| |
| |
| | |
* stable-2.14:
Update list of maintainers in pom.xml files
Change-Id: I5d63f1963eb8bca713440d37d8bd95cf20229b12
|
| |
| |
| |
| | |
Change-Id: I4b7c4abd446f243b50e21875969f7651973bbceb
|
|\|
| |
| |
| |
| |
| |
| | |
* stable-2.14:
dev-contributing.txt: Update buildifier version to 0.20.0
Change-Id: I4201ca0d3a6cc7f9be9af0e1a88823a3e6cfafb0
|
| |
| |
| |
| | |
Change-Id: I37d947d5ce0bc4be88101f9584428453dc8f492a
|
|\|
| |
| |
| |
| |
| |
| | |
* stable-2.14:
Bump minimum bazel version to 0.22.0
Change-Id: I8c8e2816aaab9df28dbfe7c0ba070a731517731d
|
| |
| |
| |
| | |
Change-Id: I4e7c826914cd4da9043fb56e28523ea6b15be9c6
|
| |
| |
| |
| |
| |
| | |
Bug: Issue 10415
Change-Id: I004b66ca39b56b801637a0a0ade6192387ed91d1
(cherry picked from commit f73f67d252debb83c7eced76ef25d016eb25c11d)
|
| |
| |
| |
| | |
Change-Id: I6c8e9a6434d128760acf2fc7c4403a917528bed8
|
| |
| |
| |
| |
| |
| |
| | |
This is a new rule as of 3.14. Which is causing the codestyle
checks to fail.
Change-Id: I408d7e5408b7ac9f88e2d72ded29b484a2aba8ad
|
| |
| |
| |
| | |
Change-Id: I5368c094a0bc3b848985fd261f474f147a945ebf
|
| |
| |
| |
| |
| |
| |
| | |
When slave mode or headless mode is activated, include '[slave]'
and '[headless]' respectively in the 'ready' info log.
Change-Id: I5789b22d32e0260cb676a91d375e919838d642a1
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Installing the online indexer in slave mode causes the daemon
startup to fail with:
No implementation for com.google.gerrit.server.index.VersionManager
was bound
when the online indexer starts.
Bug: Issue 10452
Change-Id: I0bfe06bd6e0095f94d4433e0ba95f12134da1062
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* changes:
CommitValidators: trim "ERROR" shouting from "forge committer" check
ReceiveCommits: uniformize commit validation error messages.
Inline "Change-Id" string into error messages
Always end the "Change-Id missing" error message with \n.
Print only one hint about Change-Ids at a time
Clean up Change-Id hint text
CommitValidators: Replace indexOf calls with String#contains
CommitValidators: Prefer using Splitter to String.split
|
| | |
| | |
| | |
| | | |
Change-Id: I848e210da540524142b0e44f7e826cc0d0a7646b
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Before
[abcdabc] missing subject; Change-Id must be in commit message
footer
After
commit abcdabc: missing subject; Change-Id must be in message
footer
Change-Id: I15ca67fb591b0d237cb9217936853ec8dc706aaa
|
| | |
| | |
| | |
| | | |
Change-Id: Id74f483a07dca0baf2cd50f2d26f007bdcbc8fd4
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This should fix the current output which looks like
remote: ERROR: [5bfebfb] missing .. footerHint: to automatically ..
Change-Id: Id555b98cccdc25faf2faf641f564692ced10e35d
|
| | |
| | |
| | |
| | | |
Change-Id: Idd6b5f63c71000c660a0167510d47fdd55904e92
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Before:
remote: ERROR: [c3dcb3a] ...
remote:
remote: Hint: A potential Change-Id was found, but it was not in the footer (last paragraph) of the commit message.
remote:
After:
remote: ERROR: [a1a077b] ...
remote:
remote: Hint: run
remote: git commit --amend
remote: and move 'Change-Id: Ixxx..' to the bottom on a separate line
Change-Id: I57387d88c7f105a33ede17ef76a4ac1f064fefdd
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Partial cherry-pick of commit 249d47b done in order to make
subsequent cherry-picks cleaner.
Change-Id: I7288a18d38351ebe87aa8fc25b2c232310b1273f
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Partial cherry-pick of commit 5174d2d done in order to make
subsequent cherry-picks cleaner.
Change-Id: I685ef2a3ac9ecd66aa32c6556aa1a70ba37704e1
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The current API allows to get votes either:
- By revision, but only the votes from a specific reviewer
- By change, but only all votes on the current patch set
Add a method that allows to get all the votes for a specific revision.
The result is returned as a multimap of the label name to a list
of approval infos for that label. The approval infos contain the Id,
name, username and email fields.
Change-Id: I35cde09b3ee059e8b03aff1c4ee2747944c3d37c
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The permittedVotingRange and tag parameters may be null, so
annotate them with @Nullable.
Change-Id: I6698ca34f512c783fbf319f63222804e6aec2586
|
| | |
| | |
| | |
| | | |
Change-Id: I58c0cc05409d52dabfc4f50cb21955f91e6c70a3
|
| | |
| | |
| | |
| | | |
Change-Id: I6cfe9fd05e806417740c61524e57383b93ed124d
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
cmd-review docs have been outdated for at least a year.
https://stackoverflow.com/questions/45814571/why-does-gerrit-review-return-a-not-a-valid-patch-set-error
Patch updates docs to reflect commands.
Change-Id: I33f531553a110bb92859525461dac21a02f49bd9
|