Collect response tags such as 'Reviewed-by' while parsing the stream.
This allows us to see what tags are present.
Add a new 'Fixes' tag also, since this is now quite common.
Signed-off-by: Simon Glass <sjg@chromium.org>
This patch adds support to multi-line changes. That is, if one has a line
in a changelog like
- Do a thing but
it spans multiple lines
Using Series-process-log sort would sort as if those lines were unrelated.
With this patch, any change line starting with whitespace will be
considered part of the change before it.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
By default patman generates a combined changelog for the cover letter. This
may not always be desirable.
Many patches may have the same changes. These can be coalesced with
"Series-process-log: uniq", but this is imperfect. Similar changes like
"Move foo to patch 7" will not be merged with the similar "Move foo to this
patch from patch 6".
Changes may not make sense outside of the patch they are written for. For
example, a change line of "Add check for bar" does not make sense outside
of the context in which bar might be checked for. Some changes like "New"
or "Lint" may be repeated many times throughout different change logs, but
carry no useful information in a summary.
Lastly, I like to summarize the broad strokes of the changes I have made in
the cover letter, while documenting all the details in the appropriate
patches. I think this makes it easier to get a good feel for what has
changed, without making it difficult to wade through every change in the
whole series.
This patch adds two new tags to add changelog entries which only appear in
the cover letter, or only appear in the commit. Changes documented with
"Commit-changes" will only appear in the commit, and will not appear in the
cover letter. Changes documented with "Cover-changes" will not appear in
any commit, and will only appear in the cover letter.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
At present patman sets the python path on startup so that it can access
the libraries it needs. If we convert to use absolute imports this is not
necessary.
Move patman to use absolute imports. This requires changes in tools which
use the patman libraries (which is most of them).
Signed-off-by: Simon Glass <sjg@chromium.org>
At present patman test fail in some environments which don't use utf-8
as the default file encoding. Add this explicitly.
Signed-off-by: Simon Glass <sjg@chromium.org>
As per the centithread on ksummit-discuss [1], there are folks who
feel that if a Change-Id is present in a developer's local commit that
said Change-Id could be interesting to include in upstream posts.
Specifically if two commits are posted with the same Change-Id there's
a reasonable chance that they are either the same commit or a newer
version of the same commit. Specifically this is because that's how
gerrit has trained people to work.
There is much angst about Change-Id in upstream Linux, but one thing
that seems safe and non-controversial is to include the Change-Id as
part of the string of crud that makes up a Message-Id.
Let's give that a try.
In theory (if there is enough adoption) this could help a tool more
reliably find various versions of a commit. This actually might work
pretty well for U-Boot where (I believe) quite a number of developers
use patman, so there could be critical mass (assuming that enough of
these people also use a git hook that adds Change-Id to their
commits). I was able to find this git hook by searching for "gerrit
change id git hook" in my favorite search engine.
In theory one could imagine something like this could be integrated
into other tools, possibly even git-send-email. Getting it into
patman seems like a sane first step, though.
NOTE: this patch is being posted using a patman containing this patch,
so you should be able to see the Message-Id of this patch and see that
it contains my local Change-Id, which ends in 2b9 if you want to
check.
[1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html
Signed-off-by: Douglas Anderson <dianders@chromium.org>
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
The existing test (patman --test) only covers basic checkpatch output.
We have had some problems with unicode processing and could use test
coverage for the various tags patman supports.
Add a new functional test which runs most of the patman flow on a few
test commits and checks that the results are correct.
See the documentation in the test for a description of what it does.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
There is no need for this function to return the same object that was
passed in. Drop the return value.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Unicode characters may appear in input patches so we should not warn about
them. Drop this warning.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
In python 3.x, print must be used as a function call. Convert all print
statements to the function call style, importing from __future__ where
we print with no trailing newline or print to a file object.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Simon Glass <sjg@chromium.org>
The following python error:
Traceback (most recent call last):
File "./tools/patman/patman", line 144, in <module>
series = patchstream.FixPatches(series, args)
File "./tools/patman/patchstream.py", line 477, in FixPatches
commit = series.commits[count]
IndexError: list index out of range
is seen when:
- 'END' is missing in those tags
- those tags are put in the last part in a commit message
- the commit is not the last commit of the series
Add testing logic to see if a new commit starts.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
'Series-changes' uses blank line to indicate its end. If that is
missing, series internal state variable 'in_change' may be wrong.
Correct its state.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
If 'END' is missing in a 'Cover-letter' section, and that section
happens to show up at the very end of the commit message, and the
commit is the last commit of the series, patman fails to generate
cover letter for us. Handle this in CloseCommit of patchstream.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
'Cover-letter', 'Series-notes' and 'Commit-notes' tags require an
'END' to be put at the end of its section. If we forget to put an
'END' in those sections, and these sections are followed by another
patman tag, patman generates incorrect patches. This adds codes to
handle such scenario.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Like other patman tags, use a new variable cover_match to indicate
a match for 'Cover-letter'.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
True commit lines start at column zero. Anything that is indented
is part of the commit message instead. I noticed this by trying to
run buildman with commit e3a4facdfc
as master, which contained a reference to a Linux commit inside
the commit message. ProcessLine saw that as a genuite commit
line, and thus buildman tried to build it, and died with an
exception because that SHA is not present in the U-Boot tree.
Signed-off-by: Scott Wood <scottwood@freescale.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tags like Series-version are normally expected to appear once, and with a
unique value. But buildman doesn't actually look at these tags. So ignore
conflicts.
This allows bulidman to build a branch containing multiple patman series.
Reported-by: Steve Rae <srae@broadcom.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
For reasons that are not well-understood, GetMetaDataForList() can end up
adding to an existing series even when it appears that it should be
starting a new one.
Change from using a default constructor parameter to an explicit one, to
work around this problem.
Signed-off-by: Simon Glass <sjg@chromium.org>
patman collects tags that it sees in the commit and places them nicely
sorted at the end of the patch. However, this is not really necessary and
in fact is apparently not desirable.
Suggested-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
It seems that doctest behaves differently now, and some of the unit tests
do not run. Adjust the tests to work correctly.
./tools/patman/patman --test
<unittest.result.TestResult run=10 errors=0 failures=0>
Signed-off-by: Simon Glass <sjg@chromium.org>
There is an unfortunate bug in the signoff suppression logic. The first
pass is performed with 'git log', and all signoffs are added to the
supression set, such that the second time (when processing the real
patches) we always suppress the signoffs.
Correct this by only suppressing signoffs in the second pass.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Michal Simek <monstr@monstr.eu>
Tested-by: Andreas Bießmann <andreas.devel@googlemail.com>
Add a new Patch-cc: tag which performs the service now provided by
the Cc: tag. The Cc: tag is interpreted by git send-email but
ignored by patman.
So now:
Cc: patman does nothing. (git send-email can cc patches)
Patch-cc: patman Cc's patch and removes this tag from the patch
Signed-off-by: Simon Glass <sjg@chromium.org>
Sometimes a commit should have notes enclosed with it rather
than withing the cover letter -- possibly even because there
is no cover letter. Add a 'Commit-notes' tag, similar to the
'Series-notes' one; lines between this tag and the next END
line are inserted in the patch right after the '---' commit
delimiter.
Change-Id: I01e99ae125607dc6dec08f3be8a5a0b37f0a483d
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Signed-off-by: Simon Glass <sjg@chromium.org>
(Updated README)
The git config parameter log.decorate is quite useful when working with git.
Patman, however can not handle the decorated output when parsing the commit.
To prevent this use the '--no-decorate' switch for git-log.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
Acked-by: Simon Glass <sjg@chromium.org>
For some series with lots of changes it is annoying that duplicate change
log items are not caught. It is also helpful sometimes to sort the change
logs.
Add a Series-process-log tag to enable this, which can be placed in a
commit to control this.
The change to the Cc: line is to fix a checkpatch warning.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Patman's regular expression for detecting the start of a
commit in a git log was a little simplistic and could be
confused if the git log itself had the word "commit" as
the start of a line (as this commit does). Make patman
a little more robust.
Signed-off-by: Doug Anderson <dianders@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
These tags are used by Gerrit, so let's ignore all of them.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
The cover letter is sent to everyone who is on the Cc list for any of
the patches in the series. Sometimes it is useful to send just the cover
letter to additional people, so that they are aware of the series, but
don't need to wade through all the individual patches.
Add a new Cover-letter-cc tag for this purpose.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Although "Reviewed-by:" is a tag that gerrit adds, it's also a tag
used by upstream. Stripping it is undesirable. In fact, we should
treat it as important.
Signed-off-by: Doug Anderson <dianders@chromium.org>
Reviewed-by: Otavio Salvador <otavio@ossystems.com.br>
Acked-by: Simon Glass <sjg@chromium.org>
We normally read from the current branch, but buildman will need to look
at commits from another branch. Allow the metadata to be read from any
list of commits, to provide this flexibility.
Signed-off-by: Simon Glass <sjg@chromium.org>
Rather than returning a list of things, return an object. That makes it
easier to access the returned items, and easier to extend the return
value later.
Signed-off-by: Simon Glass <sjg@chromium.org>
The BRANCH= tag can be used to indicate the destination branch for a
commit. Ignore this tag.
Also ignore the gerrit 'Commit-Ready:' tag.
Signed-off-by: Simon Glass <sjg@chromium.org>
Colored logs confuse patman when analyzing logs.
Add --no-color option in git log commands in case
the default config has color.
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@ti.com>
Currently patman assumes that there should be only one Signoff line
and this is obviously incorrect: we often have to work with patches
containing other people signoffs. Moreover, it's really desirable
to preserve the comments between signoffs.
So until some sophisticated signoff processing will be developed I
suggest just don't mess with signoffs at all and treat them like
plain text lines. The only drawback I've found so far is the case
where you have a patch with someones else signoff but not yours and
also have to patman tags under signoff line. In this case you will
get extra empty line between signoffs.
Signed-off-by: Ilya Yanok <ilya.yanok@cogentembedded.com>
We already got all changes from git log output and the comment
to the ProcessLine function clearly states that 'patch' mode
is not for scanning tags.
Signed-off-by: Ilya Yanok <ilya.yanok@cogentembedded.com>
Changes may end in '---' line or Signoff line (generated by
git format-patch) in case of Series-changes: lines being
the last ones in commit message. So detect it properly.
Signed-off-by: Ilya Yanok <ilya.yanok@cogentembedded.com>
What is this?
=============
This tool is a Python script which:
- Creates patch directly from your branch
- Cleans them up by removing unwanted tags
- Inserts a cover letter with change lists
- Runs the patches through checkpatch.pl and its own checks
- Optionally emails them out to selected people
It is intended to automate patch creation and make it a less
error-prone process. It is useful for U-Boot and Linux work so far,
since it uses the checkpatch.pl script.
It is configured almost entirely by tags it finds in your commits.
This means that you can work on a number of different branches at
once, and keep the settings with each branch rather than having to
git format-patch, git send-email, etc. with the correct parameters
each time. So for example if you put:
in one of your commits, the series will be sent there.
See the README file for full details.
END
Signed-off-by: Simon Glass <sjg@chromium.org>