* np/pack:
fix repack with --max-pack-size
builtin-pack-object: cache small deltas
git-pack-objects: cache small deltas between big objects
builtin-pack-objects: don't fail, if delta is not possible
* sp/pack:
Style nit - don't put space after function names
Ensure the pack index is opened before access
Simplify index access condition in count-objects, pack-redundant
Test for recent rev-parse $abbrev_sha1 regression
rev-parse: Identify short sha1 sums correctly.
Attempt to delay prepare_alt_odb during get_sha1
Micro-optimize prepare_alt_odb
Lazily open pack index files on demand
The command is part of the main porcelain making git-add more
appropriate.
Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
* maint:
Use =20 when rfc2047 encoding spaces.
Create a new manpage for the gitignore format, and reference it elsewhere
Documentation: robustify asciidoc GIT_VERSION replacement
Encode ' ' using '=20' even though rfc2047 allows using '_' for
readability. Unfortunately, many programs do not understand this and
just leave the underscore in place. Using '=20' seems to work better.
[jc: with adjustment to t3901]
Signed-off-by: Kristian Høgsberg <hoegsberg@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Only git-ls-files(1) describes the gitignore format in detail, and it does so
with reference to git-ls-files options. Most users don't use the plumbing
command git-ls-files directly, and shouldn't have to look in its manpage for
information on the gitignore format.
Create a new manpage gitignore(5) (Documentation/gitignore.txt), and factor
out the gitignore documentation into that file, changing it to refer to
.gitignore and $GIT_DIR/info/exclude as used by porcelain commands. Reference
gitignore(5) from other relevant manpages and documentation. Remove
now-redundant information on exclude patterns from git-ls-files(1), leaving
only information on how git-ls-files options specify exclude patterns and what
precedence they have.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Instead of using sed on the resulting file, we now have a
git_version asciidoc attribute. This means that we don't
pipe the output of asciidoc, which means we can detect build
failures.
Problem reported by Scott Lamb, solution suggested by Jonas Fonseca.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Its wrong to exit the application if we destroy a random widget
contained withing something else; especially if its some small
trivial thing that has no impact on the overall operation.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
To improve performance on fork+exec impoverished systems (such as
Windows) we want to avoid running git-symbolic-ref on every rescan
if we can do so. A quick way to implement such an avoidance is to
just read the HEAD ref ourselves; we'll either see it as a symref
(starts with "ref: ") or we'll see it as a detached head (40 hex
digits). In either case we can treat that as our current branch.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Johannes Sixt pointed out that dropping to 0 lines of context
does allow the user to get more fine-grained hunk selection,
especially since we don't currently support "highlight and
apply (or revert)".
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Commit 83572c1a91 changed many
realloc to xrealloc. This change was made in diff-delta.c too,
although the code can handle an out of memory failure.
This patch reverts this change in diff-delta.c.
Signed-off-by: Martin Koegler <mkoegler@auto.tuwien.ac.at>
Signed-off-by: Junio C Hamano <junkio@cox.net>
* maint:
git-config: Improve documentation of git-config file handling
git-config: Various small fixes to asciidoc documentation
decode_85(): fix missing return.
fix signed range problems with hex conversions
* maint-1.5.1:
git-config: Improve documentation of git-config file handling
git-config: Various small fixes to asciidoc documentation
decode_85(): fix missing return.
fix signed range problems with hex conversions
The description which files git-config uses and how the various
command line options and environment variables affect its
behaviour was incomplete, outdated and confusing.
Signed-off-by: Frank Lichtenheld <frank@lichtenheld.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Add '' around the only mentioned commandline option that didn't
have it.
Make reference to section EXAMPLE a link and rename it to
EXAMPLES because it actually contains a lot of examples.
Signed-off-by: Frank Lichtenheld <frank@lichtenheld.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Jon Smirl said:
| Once an object reference hits a pack file it is very likely that
| following references will hit the same pack file. So first place to
| look for an object is the same place the previous object was found.
This is indeed a good heuristic so here it is. The search always start
with the pack where the last object lookup succeeded. If the wanted
object is not available there then the search continues with the normal
pack ordering.
To test this I split the Linux repository into 66 packs and performed a
"time git-rev-list --objects --all > /dev/null". Best results are as
follows:
Pack Sort w/o this patch w/ this patch
-------------------------------------------------------------
recent objects last 26.4s 20.9s
recent objects first 24.9s 18.4s
This shows that the pack order based on object age has some influence,
but that the last-used-pack heuristic is even more significant in
reducing object lookup.
Signed-off-by: Nicolas Pitre <nico@cam.org> --- Note: the
--max-pack-size to git-repack currently produces packs with old objects
after those containing recent objects. The pack sort based on
filesystem timestamp is therefore backward for those. This needs to be
fixed of course, but at least it made me think about this variable for
the test.
Signed-off-by: Junio C Hamano <junkio@cox.net>
Two issues here:
1) git-repack -a --max-pack-size=10 on the GIT repo dies pretty quick.
There is a lot of confusion about deltas that were suposed to be
reused from another pack but that get stored undeltified due to pack
limit and object size doesn't match entry->size anymore. This test
is not really worth the complexity for determining when it is valid
so get rid of it.
2) If pack limit is reached, the object buffer is freed, including when
it comes from a cached delta data. In practice the object will be
stored in a subsequent pack undeltified, but let's make sure no
pointer to freed data subsists by clearing entry->delta_data.
I also reorganized that code a bit to make it more readable.
Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Use a generic make rule to build all the test programs, rather than
specifically mentioning each one.
Signed-off-by: Julian Phillips <julian@quantumfyre.co.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
If the user has no branches at all (their refs/heads/ is empty)
and they are on a detached HEAD we have a valid repository but
there are no branches to populate into the branch pulldown in
the create branch dialog. Instead of erroring out we can skip
that part of the dialog, much like we do with tracking branches
or tags when the user doesn't have any.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When the function detected an invalid base85 sequence, it issued
an error message but forgot to return error status at that point
and kept going.
Signed-off-by: Jerald Fitzjerald <jfj@freemail.gr>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Make hexval_table[] "const". Also make sure that the accessor
function hexval() does not access the table with out-of-range
values by declaring its parameter "unsigned char", instead of
"unsigned int".
With this, gcc can just generate:
movzbl (%rdi), %eax
movsbl hexval_table(%rax),%edx
movzbl 1(%rdi), %eax
movsbl hexval_table(%rax),%eax
sall $4, %edx
orl %eax, %edx
for the code to generate a byte from two hex characters.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Our style is to not put a space after a function name. I did here,
and Junio applied the patch with the incorrect formatting. So I'm
cleaning up after myself since I noticed it upon review.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
In this particular location of fsck the index should have already
been opened by verify_pack, which is called just before we get
here and loop through the object names. However, just in case a
future version of that function does not use the index file we'll
double-check its open before we access the num_objects field.
Better safe now than sorry later.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
My earlier lazy index opening patch changed this condition to check
index_data and call open_pack_index if it was NULL. In truth we only
care about num_objects. Since open_pack_index does no harm if the
index is already open, and all indexes are likely to be closed in
this application, the "performance optimization" of inlining the
index_data check here was wrong.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
My recent patch "Lazily open pack index files on demand" caused a
regression in the case of parsing abbreviated SHA-1 object names.
Git was unable to translate the abbreviated name into the full name
if the object was packed, as the pack .idx files were not opened
before being accessed.
This is a simple test to repack a repository then test for an
abbreviated SHA-1 within the packfile.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
find_short_packed_object was not loading the pack index files.
Teach it to do so.
Signed-off-by: James Bowes <jbowes@dangerouslyinc.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Creating deltas between big blobs is a CPU and memory intensive task.
In the writing phase, all (not reused) deltas are redone.
This patch adds support for caching deltas from the deltifing phase, so
that that the writing phase is faster.
The caching is limited to small deltas to avoid increasing memory usage very much.
The implemented limit is (memory needed to create the delta)/1024.
Signed-off-by: Martin Koegler <mkoegler@auto.tuwien.ac.at>
Signed-off-by: Junio C Hamano <junkio@cox.net>
If builtin-pack-objects runs out of memory while finding
the best deltas, it bails out with an error.
If the delta index creation fails (because there is not enough memory),
we can downgrade the error message to a warning and continue with the
next object.
Signed-off-by: Martin Koegler <mkoegler@auto.tuwien.ac.at>
Signed-off-by: Junio C Hamano <junkio@cox.net>
* db/remote:
Move refspec pattern matching to match_refs().
Update local tracking refs when pushing
Add handlers for fetch-side configuration of remotes.
Move refspec parser from connect.c and cache.h to remote.{c,h}
Move remote parsing into a library file out of builtin-push.
* jc/nodelta:
builtin-pack-objects: remove unnecessary code for no-delta
Teach "delta" attribute to pack-objects.
pack-objects: pass fullname down to add_object_entry()
* ar/run:
Allow environment variables to be unset in the processes started by run_command
Add ability to specify environment extension to run_command
Add run_command_v_opt_cd: chdir into a directory before exec
* maint:
cvsserver: Fix some typos in asciidoc documentation
cvsserver: Note that CVS_SERVER can also be specified as method variable
cvsserver: Correct inetd.conf example in asciidoc documentation
user-manual: fixed typo in example
Add test case for $Id$ expanded in the repository
git-svn: avoid md5 calculation entirely if SVN doesn't provide one
Makefile: Remove git-fsck and git-verify-pack from PROGRAMS
Fix stupid typo in lookup_tag()
git-gui: Guess our share/git-gui/lib path at runtime if possible
Correct key bindings to Control-<foo>
git-gui: Tighten internal pattern match for lib/ directory
Since this is a trivial variation of the general pserver
authentication, there is really no reason not to support
it.
Signed-off-by: Frank Lichtenheld <frank@lichtenheld.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
If we are too fast with our changes, the file in
the working copy might still have the same mtime
as noted in the CVS/Entries. This will cause CVS
to happily report to the server that the file is
unmodified which can lead to data loss (and in
our case test failure).
CVS sucks!
Work around that by sleeping for a second.
Signed-off-by: Frank Lichtenheld <frank@lichtenheld.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
On one particular system I like to keep a cluster of bare Git
repositories and spawn new-workdirs off of them. Since the bare
repositories don't have working directories associated with them
they don't have a .git/ subdirectory that hosts the repository we
are linking to.
Using a bare repository as the backing repository for a workdir
created by this script does require that the user delete core.bare
from the repository's configuration file, so that Git auto-senses
the bareness of a repository based on pathname information, and
not based on the config file.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Earlier commit d63bd9a broke the case where more than one input
files are fed to mailsplit by not incrementing the base counter
when splitting second and subsequent input files. This should
fix it.
Signed-off-by: Junio C Hamano <junkio@cox.net>
