git/commit.c, branch v2.0.3

Merge branch 'bg/xcalloc-nmemb-then-size' into maint

2014-07-22T17:25:17Z

* bg/xcalloc-nmemb-then-size: transport-helper.c: rearrange xcalloc arguments remote.c: rearrange xcalloc arguments reflog-walk.c: rearrange xcalloc arguments pack-revindex.c: rearrange xcalloc arguments notes.c: rearrange xcalloc arguments imap-send.c: rearrange xcalloc arguments http-push.c: rearrange xcalloc arguments diff.c: rearrange xcalloc arguments config.c: rearrange xcalloc arguments commit.c: rearrange xcalloc arguments builtin/remote.c: rearrange xcalloc arguments builtin/ls-remote.c: rearrange xcalloc arguments

reuse cached commit buffer when parsing signatures

2014-06-13T19:10:13Z

When we call show_signature or show_mergetag, we read the commit object fresh via read_sha1_file and reparse its headers. However, in most cases we already have the object data available, attached to the "struct commit". This is partially laziness in dealing with the memory allocation issues, but partially defensive programming, in that we would always want to verify a clean version of the buffer (not one that might have been munged by other users of the commit). However, we do not currently ever munge the commit buffer, and not using the already-available buffer carries a fairly big performance penalty when we are looking at a large number of commits. Here are timings on linux.git: [baseline, no signatures] $ time git log >/dev/null real 0m4.902s user 0m4.784s sys 0m0.120s [before] $ time git log --show-signature >/dev/null real 0m14.735s user 0m9.964s sys 0m0.944s [after] $ time git log --show-signature >/dev/null real 0m9.981s user 0m5.260s sys 0m0.936s Note that our user CPU time drops almost in half, close to the non-signature case, but we do still spend more wall-clock and system time, presumably from dealing with gpg. An alternative to this is to note that most commits do not have signatures (less than 1% in this repo), yet we pay the re-parsing cost for every commit just to find out if it has a mergetag or signature. If we checked that when parsing the commit initially, we could avoid re-examining most commits later on. Even if we did pursue that direction, however, this would still speed up the cases where we _do_ have signatures. So it's probably worth doing either way. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

commit: record buffer length in cache

2014-06-13T19:09:38Z

Most callsites which use the commit buffer try to use the cached version attached to the commit, rather than re-reading from disk. Unfortunately, that interface provides only a pointer to the NUL-terminated buffer, with no indication of the original length. For the most part, this doesn't matter. People do not put NULs in their commit messages, and the log code is happy to treat it all as a NUL-terminated string. However, some code paths do care. For example, when checking signatures, we want to be very careful that we verify all the bytes to avoid malicious trickery. This patch just adds an optional "size" out-pointer to get_commit_buffer and friends. The existing callers all pass NULL (there did not seem to be any obvious sites where we could avoid an immediate strlen() call, though perhaps with some further refactoring we could). Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

commit: convert commit->buffer to a slab

2014-06-13T19:08:17Z

This will make it easier to manage the buffer cache independently of the "struct commit" objects. It also shrinks "struct commit" by one pointer, which may be helpful. Unfortunately it does not reduce the max memory size of something like "rev-list", because rev-list uses get_cached_commit_buffer() to decide not to show each commit's output (and due to the design of slab_at, accessing the slab requires us to extend it, allocating exactly the same number of buffer pointers we dropped from the commit structs). Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

use get_commit_buffer to avoid duplicate code

2014-06-13T19:08:17Z

For both of these sites, we already do the "fallback to read_sha1_file" trick. But we can shorten the code by just using get_commit_buffer. Note that the error cases are slightly different when read_sha1_file fails. get_commit_buffer will die() if the object cannot be loaded, or is a non-commit. For get_sha1_oneline, this will almost certainly never happen, as we will have just called parse_object (and if it does, it's probably worth complaining about). For record_author_date, the new behavior is probably better; we notify the user of the error instead of silently ignoring it. And because it's used only for sorting by author-date, somebody examining a corrupt repo can fallback to the regular traversal order. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

provide helpers to access the commit buffer

2014-06-13T19:08:17Z

Many sites look at commit->buffer to get more detailed information than what is in the parsed commit struct. However, we sometimes drop commit->buffer to save memory, in which case the caller would need to read the object afresh. Some callers do this (leading to duplicated code), and others do not (which opens the possibility of a segfault if somebody else frees the buffer). Let's provide a pair of helpers, "get" and "unuse", that let callers easily get the buffer. They will use the cached buffer when possible, and otherwise load from disk using read_sha1_file. Note that we also need to add a "get_cached" variant which returns NULL when we do not have a cached buffer. At first glance this seems to defeat the purpose of "get", which is to always provide a return value. However, some log code paths actually use the NULL-ness of commit->buffer as a boolean flag to decide whether to try printing the commit. At least for now, we want to continue supporting that use. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

provide a helper to set the commit buffer

2014-06-13T19:08:17Z

Right now this is just a one-liner, but abstracting it will make it easier to change later. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

provide a helper to free commit buffer

2014-06-13T19:07:47Z

This converts two lines into one at each caller. But more importantly, it abstracts the concept of freeing the buffer, which will make it easier to change later. Note that we also need to provide a "detach" mechanism for a tricky case in index-pack. We are passed a buffer for the object generated by processing the incoming pack. If we are not using --strict, we just calculate the sha1 on that buffer and return, leaving the caller to free it. But if we are using --strict, we actually attach that buffer to an object, pass the object to the fsck functions, and then detach the buffer from the object again (so that the caller can free it as usual). In this case, we don't want to free the buffer ourselves, but just make sure it is no longer associated with the commit. Note that we are making the assumption here that the attach/detach process does not impact the buffer at all (e.g., it is never reallocated or modified). That holds true now, and we have no plans to change that. However, as we abstract the commit_buffer code, this dependency becomes less obvious. So when we detach, let's also make sure that we get back the same buffer that we gave to the commit_buffer code. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

commit: push commit_index update into alloc_commit_node

2014-06-12T17:29:42Z

Whenever we create a commit object via lookup_commit, we give it a unique index to be used with the commit-slab API. The theory is that any "struct commit" we create would follow this code path, so any such struct would get an index. However, callers could use alloc_commit_node() directly (and get multiple commits with index 0). Let's push the indexing into alloc_commit_node so that it's hard for callers to get it wrong. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

commit_tree: take a pointer/len pair rather than a const strbuf

2014-06-12T17:29:41Z

While strbufs are pretty common throughout our code, it is more flexible for functions to take a pointer/len pair than a strbuf. It's easy to turn a strbuf into such a pair (by dereferencing its members), but less easy to go the other way (you can strbuf_attach, but that has implications about memory ownership). This patch teaches commit_tree (and its associated callers and sub-functions) to take such a pair for the commit message rather than a strbuf. This makes passing the buffer around slightly more verbose, but means we can get rid of some dangerous strbuf_attach calls in the next patch. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano