Commit Briefs
fix an issue where 'git fetch' would error or hang against gotd
If Git has more than 16 have-lines to send it will send a flush-pkt followed by more have-lines. Due to a misunderstanding on my part, gotd didn't like this because it assumed that the flush-pkt terminates the list. Add a test coverage in a new file which we can use to test Git interop issues. Fixes a problem seen by Thomas Adam upon git fetch from got.g.o.
remove unneded wbuf->fd = -1
There's no need to set the fd to -1 on ibufs created with imsg_create(3), and it was probably never needed.
convert to use imsg_get_fd()
While here also fix a fd leak in got-read-pack. We were dup'ing imsg.fd without closing imsg.fd later; instead just use imsg_get_fd() to extract the file descriptor. Tested by falsifian and Kyle Ackerman, thanks! 'go ahead' stsp@
make gotsh handle poll timeouts when checking for gotd closing the pack pipe
This will hopefully fix issues seen on servers where gotsh exits early for no good reason.
gotsh: do not set POLLOUT flag when there is no data to send
Prevent an endless polling loop on Linux where POLLOUT and POLLHUP can both be reported via pollfd revents.
make gotd avoid a "failed to push some refs" error from no-op 'git push'
gotsh was sending errors such as "unexpected end of file" and "unexpected flush-pkt" to the client when the client disconnected immediately after receiving reference announcements from the server. As a result, 'git push' with both sides up-to-date would show an obscure error message: = [up to date] main -> main error: failed to push some refs to 'ssh://example.com/git/repo.git Now we see: = [up to date] main -> main Everything up-to-date ok jamsek
significantly reduce the amount of code linked into gitwrapper
By moving got_serve_parse_command() from lib/serve.c into lib/dial.c as got_dial_parse_command(), we can significantly reduce the amount of symbols gitwrapper depends on indirectly. As a downside, gotsh now needs to link to dial.c. But it only uses the same parsing routine, and any other routines in dial.c would likely cause pledge violations in gotsh if used. No functional change.
fix gotd sending too large pack files in some cases
Make gotsh pass all have-lines sent by clients onwards to gotd, reducing size of pack files generated by the repo_read process. Problem reported by naddy@ ok jamsek
provide functions to parse/serialize different hashes
it abstracts over the hash type and ensures that object ids are zero'ed before their sha1 digest is written. Needed by the incoming sha256 support. ok stsp@
rename lib/sha1.c to lib/hash.c
It will soon grow functions to deal with sha256 too. stsp@ agrees.
include sha2.h too where sha1.h is included
In preparation for wide sha256 support; stsp@ agrees. Change done mechanically with find . -iname \*.[cy] -exec sam {} + X ,x/<sha1\.h>/i/\n#include <sha2.h>
gotd: implement the delete-refs capability
Allow clients to run "got send -d" against gotd. Clients will send a zero-id as new id for a reference and, in the special but more common case of just deleting and not updating, no pack file will be sent. ok and tweaks by stsp@
gotsh: validate with parse_command before connecting
Export parse_command (now got_serve_parse_command) from lib/serve.c and use it to validate the command line in gotsh instead of just checking that the -c argument starts with 'git-receive-pack' or 'git-upload-pack'. Invalid usage now always fails before opening the socket, while before it wasn't always the case. This also means that invalid usage doesn't count towards the limits. ok jamsek, stsp