When debugging the pack protocol, it is sometimes useful to
store the verbatim pack that we sent or received on the
wire. Looking at the on-disk result is often not helpful for
a few reasons:
1. If the operation is a clone, we destroy the repo on
failure, leaving nothing on disk.
2. If the pack is small, we unpack it immediately, and the
full pack never hits the disk.
3. If we feed the pack to "index-pack --fix-thin", the
resulting pack has the extra delta bases added to it.
We already have a GIT_TRACE_PACKET mechanism for tracing
packets. Let's extend it with GIT_TRACE_PACKFILE to dump the
verbatim packfile.
There are a few other positive fallouts that come from
rearranging this code:
- We currently disable the packet trace after seeing the
PACK header, even though we may get human-readable lines
on other sidebands; now we include them in the trace.
- We currently try to print "PACK ..." in the trace to
indicate that the packfile has started. But because we
disable packet tracing, we never printed this line. We
will now do so.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
To find the start of the pack data, we accept the word PACK
at the beginning of any sideband channel, even though what
we really want is to find the pack data on channel 1. In
practice this doesn't matter, as sideband-2 messages tend to
start with "error:" or similar, but it is a good idea to be
explicit (especially as we add more code in this area, we
will rely on this assumption).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We carefully check that our pkt buffer has enough characters
before seeing if it starts with "PACK". The intent is to
avoid reading random memory if we get a short buffer like
"PAC".
However, we know that the traced packets are always
NUL-terminated. They come from one of these sources:
1. A string literal.
2. `format_packet`, which uses a strbuf.
3. `packet_read`, which defensively NUL-terminates what we
read.
We can therefore drop the length checks, as we know we will
hit the trailing NUL if we have a short input.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we send out pkt-lines with refnames, we use a static
1000-byte buffer. This means that the maximum size of a ref
over the git protocol is around 950 bytes (the exact size
depends on the protocol line being written, but figure on a sha1
plus some boilerplate).
This is enough for any sane workflow, but occasionally odd
things happen (e.g., a bug may create a ref "foo/foo/foo/..."
accidentally). With the current code, you cannot even use
"push" to delete such a ref from a remote.
Let's switch to using a strbuf, with a hard-limit of
LARGE_PACKET_MAX (which is specified by the protocol). This
matches the size of the readers, as of 74543a0 (pkt-line:
provide a LARGE_PACKET_MAX static buffer, 2013-02-20).
Versions of git older than that will complain about our
large packets, but it's really no worse than the current
behavior. Right now the sender barfs with "impossibly long
line" trying to send the packet, and afterwards the reader
will barf with "protocol error: bad line length %d", which
is arguably better anyway.
Note that we're not really _solving_ the problem here, but
just bumping the limits. In theory, the length of a ref is
unbounded, and pkt-line can only represent sizes up to
65531 bytes. So we are just bumping the limit, not removing
it. But hopefully 64K should be enough for anyone.
As a bonus, by using a strbuf for the formatting we can
eliminate an unnecessary copy in format_buf_write.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The trace API currently rechecks the environment variable and reopens the
trace file on every API call. This has the ugly side effect that errors
(e.g. file cannot be opened, or the user specified a relative path) are
also reported on every call. Performance can be improved by about factor
three by remembering the environment state and keeping the file open.
Replace the 'const char *key' parameter in the API with a pointer to a
'struct trace_key' that bundles the environment variable name with
additional, trace-internal state. Change the call sites of these APIs to
use a static 'struct trace_key' instead of a string constant.
In trace.c::get_trace_fd(), save and reuse the file descriptor in 'struct
trace_key'.
Add a 'trace_disable()' API, so that packet_trace() can cleanly disable
tracing when it encounters packed data (instead of using unsetenv()).
Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Leaving only the function definitions and declarations so that any
new topic in flight can still make use of the old functions, replace
existing uses of the prefixcmp() and suffixcmp() with new API
functions.
The change can be recreated by mechanically applying this:
$ git grep -l -e prefixcmp -e suffixcmp -- \*.c |
grep -v strbuf\\.c |
xargs perl -pi -e '
s|!prefixcmp\(|starts_with\(|g;
s|prefixcmp\(|!starts_with\(|g;
s|!suffixcmp\(|ends_with\(|g;
s|suffixcmp\(|!ends_with\(|g;
'
on the result of preparatory changes in this series.
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The packet_read function reads from a descriptor. The
packet_get_line function is similar, but reads from an
in-memory buffer, and uses a completely separate
implementation. This patch teaches the generic packet_read
function to accept either source, and we can do away with
packet_get_line's implementation.
There are two other differences to account for between the
old and new functions. The first is that we used to read
into a strbuf, but now read into a fixed size buffer. The
only two callers are fine with that, and in fact it
simplifies their code, since they can use the same
static-buffer interface as the rest of the packet_read_line
callers (and we provide a similar convenience wrapper for
reading from a buffer rather than a descriptor).
This is technically an externally-visible behavior change in
that we used to accept arbitrary sized packets up to 65532
bytes, and now cap out at LARGE_PACKET_MAX, 65520. In
practice this doesn't matter, as we use it only for parsing
smart-http headers (of which there is exactly one defined,
and it is small and fixed-size). And any extension headers
would be breaking the protocol to go over LARGE_PACKET_MAX
anyway.
The other difference is that packet_get_line would return
on error rather than dying. However, both callers of
packet_get_line are actually improved by dying.
The first caller does its own error checking, but we can
drop that; as a result, we'll actually get more specific
reporting about protocol breakage when packet_read dies
internally. The only downside is that packet_read will not
print the smart-http URL that failed, but that's not a big
deal; anybody not debugging can already see the remote's URL
already, and anybody debugging would want to run with
GIT_CURL_VERBOSE anyway to see way more information.
The second caller, which is just trying to skip past any
extra smart-http headers (of which there are none defined,
but which we allow to keep room for future expansion), did
not error check at all. As a result, it would treat an error
just like a flush packet. The resulting mess would generally
cause an error later in get_remote_heads, but now we get
error reporting much closer to the source of the problem.
Brown-paper-bag-fixes-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Most of the callers of packet_read_line just read into a
static 1000-byte buffer (callers which handle arbitrary
binary data already use LARGE_PACKET_MAX). This works fine
in practice, because:
1. The only variable-sized data in these lines is a ref
name, and refs tend to be a lot shorter than 1000
characters.
2. When sending ref lines, git-core always limits itself
to 1000 byte packets.
However, the only limit given in the protocol specification
in Documentation/technical/protocol-common.txt is
LARGE_PACKET_MAX; the 1000 byte limit is mentioned only in
pack-protocol.txt, and then only describing what we write,
not as a specific limit for readers.
This patch lets us bump the 1000-byte limit to
LARGE_PACKET_MAX. Even though git-core will never write a
packet where this makes a difference, there are two good
reasons to do this:
1. Other git implementations may have followed
protocol-common.txt and used a larger maximum size. We
don't bump into it in practice because it would involve
very long ref names.
2. We may want to increase the 1000-byte limit one day.
Since packets are transferred before any capabilities,
it's difficult to do this in a backwards-compatible
way. But if we bump the size of buffer the readers can
handle, eventually older versions of git will be
obsolete enough that we can justify bumping the
writers, as well. We don't have plans to do this
anytime soon, but there is no reason not to start the
clock ticking now.
Just bumping all of the reading bufs to LARGE_PACKET_MAX
would waste memory. Instead, since most readers just read
into a temporary buffer anyway, let's provide a single
static buffer that all callers can use. We can further wrap
this detail away by having the packet_read_line wrapper just
use the buffer transparently and return a pointer to the
static storage. That covers most of the cases, and the
remaining ones already read into their own LARGE_PACKET_MAX
buffers.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The packets sent during ref negotiation are all terminated
by newline; even though the code to chomp these newlines is
short, we end up doing it in a lot of places.
This patch teaches packet_read_line to auto-chomp the
trailing newline; this lets us get rid of a lot of inline
chomping code.
As a result, some call-sites which are not reading
line-oriented data (e.g., when reading chunks of packfiles
alongside sideband) transition away from packet_read_line to
the generic packet_read interface. This patch converts all
of the existing callsites.
Since the function signature of packet_read_line does not
change (but its behavior does), there is a possibility of
new callsites being introduced in later commits, silently
introducing an incompatibility. However, since a later
patch in this series will change the signature, such a
commit would have to be merged directly into this commit,
not to the tip of the series; we can therefore ignore the
issue.
This is an internal cleanup and should produce no change of
behavior in the normal case. However, there is one corner
case to note. Callers of packet_read_line have never been
able to tell the difference between a flush packet ("0000")
and an empty packet ("0004"), as both cause packet_read_line
to return a length of 0. Readers treat them identically,
even though Documentation/technical/protocol-common.txt says
we must not; it also says that implementations should not
send an empty pkt-line.
By stripping out the newline before the result gets to the
caller, we will now treat the newline-only packet ("0005\n")
the same as an empty packet, which in turn gets treated like
a flush packet. In practice this doesn't matter, as neither
empty nor newline-only packets are part of git's protocols
(at least not for the line-oriented bits, and readers who
are not expecting line-oriented packets will be calling
packet_read directly, anyway). But even if we do decide to
care about the distinction later, it is orthogonal to this
patch. The right place to tighten would be to stop treating
empty packets as flush packets, and this change does not
make doing so any harder.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Originally we had a single function for reading packetized
data: packet_read_line. Commit 46284dd grew a more "gentle"
form, packet_read, that returns an error instead of dying
upon reading a truncated input stream. However, it is not
clear from the names which should be called, or what the
difference is.
Let's instead make packet_read be a generic public interface
that can take option flags, and update the single callsite
that uses it. This is less code, more clear, and paves the
way for introducing more options into the generic interface
later. The function signature is changed, so there should be
no hidden conflicts with topics in flight.
While we're at it, we'll document how error conditions are
handled based on the options, and rename the confusing
"return_line_fail" option to "gentle_on_eof". While we are
cleaning up the names, we can drop the "return_line_fail"
checks in packet_read_internal entirely. They look like
this:
ret = safe_read(..., return_line_fail);
if (return_line_fail && ret < 0)
...
The check for return_line_fail is a no-op; safe_read will
only ever return an error value if return_line_fail was true
in the first place.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is just write_or_die by another name. The one
distinction is that write_or_die will treat EPIPE specially
by suppressing error messages. That's fine, as we die by
SIGPIPE anyway (and in the off chance that it is disabled,
write_or_die will simulate it).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The comment describing the packet writing interface was
originally written above packet_write, but migrated to be
above safe_write in f3a3214, probably because it is meant to
generally describe the packet writing interface and not a
single function. Let's move it into the header file, where
users of the interface are more likely to see it.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If a server accessed through ssh is denying access git will currently
issue the message
"fatal: The remote end hung up unexpectedly"
as the last line. This sounds as if something really ugly just happened.
Since this is a quite typical situation in which users regularly get
we do not say that if it happens at the beginning when reading the
remote heads.
If its in the very first beginning of reading the remote heads it is
very likely an authentication error or a missing repository.
If it happens later during reading the remote heads we still indicate
that it happened during this initial contact phase.
Signed-off-by: Heiko Voigt <hvoigt@hvoigt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* load_file() returns a void pointer but is using 0 for the return
value
* builtin/receive-pack.c forgot to include builtin.h
* packet_trace_prefix can be marked static
* ll_merge takes a pointer for its last argument, not an int
* crc32 expects a pointer as the second argument but Z_NULL is defined
to be 0 (see 38f4d13 sparse fix: Using plain integer as NULL pointer,
2006-11-18 for more info)
Signed-off-by: Stephen Boyd <bebarino@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This shows a trace of all packets coming in or out of a given
program. This can help with debugging object negotiation or
other protocol issues.
To keep the code changes simple, we operate at the lowest
level, meaning we don't necessarily understand what's in the
packets. The one exception is a packet starting with "PACK",
which causes us to skip that packet and turn off tracing
(since the gigantic pack data will not be interesting to
read, at least not in the trace format).
We show both written and read packets. In the local case,
this may mean you will see packets twice (written by the
sender and read by the receiver). However, for cases where
the other end is remote, this allows you to see the full
conversation.
Packet tracing can be enabled with GIT_TRACE_PACKET=<foo>,
where <foo> takes the same arguments as GIT_TRACE.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When there is an error parsing the 4 byte length component we now
display it as part of the die message, this may hint as to what
data was misunderstood by the application.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
These routines help to work with pkt-line values inside of a strbuf,
permitting simple formatting of buffered network messages.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change calls to die(..., strerror(errno)) to use the new die_errno().
In the process, also make slight style adjustments: at least state
_something_ about the function that failed (instead of just printing
the pathname), and put paths in single quotes.
Signed-off-by: Thomas Rast <trast@student.ethz.ch>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This uses "git-apply --whitespace=strip" to fix whitespace errors that have
crept in to our source files over time. There are a few files that need
to have trailing whitespaces (most notably, test vectors). The results
still passes the test, and build result in Documentation/ area is unchanged.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This patch fixes all calls to xread() where the return value is not
stored into an ssize_t. The patch should not have any effect whatsoever,
other than putting better/more appropriate type names on variables.
Signed-off-by: Johan Herland <johan@herland.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
The remote server might not want to tell why it doesn't like us for
security reasons, but let's make the client report such error in a bit
less confusing way. The remote failure remains a mystery, but the local
message might be a bit less so.
[jc: with a gentle wording updates from Andy Parkins]
Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
This implements a protocol extension between fetch-pack and
upload-pack to allow stderr stream from upload-pack (primarily
used for the progress bar display) to be passed back.
Signed-off-by: Junio C Hamano <junkio@cox.net>
ANSI C99 doesn't allow void-pointer arithmetic. This patch fixes this in
various ways. Usually the strategy that required the least changes was used.
Signed-off-by: Florian Forster <octo@verplant.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
We had errno==EINTR check after read(2)/write(2) sprinkled all
over the places, always doing continue. Consolidate them into
xread()/xwrite() wrapper routines.
Credits for suggestion goes to HPA -- bugs are mine.
Signed-off-by: Junio C Hamano <junkio@cox.net>
Jeff King
Karsten Blees
Christian Couder
Heiko Voigt
Stephen Boyd
Shawn O. Pearce
Thomas Rast
Heikki Orsila
Junio C Hamano
Johan Herland
Nicolas Pitre
Junio C Hamano
Florian Forster
Linus Torvalds