* jk/skip-prefix:
http-push: refactor parsing of remote object names
imap-send: use skip_prefix instead of using magic numbers
use skip_prefix to avoid repeated calculations
git: avoid magic number with skip_prefix
fetch-pack: refactor parsing in get_ack
fast-import: refactor parsing of spaces
stat_opt: check extra strlen call
daemon: use skip_prefix to avoid magic numbers
fast-import: use skip_prefix for parsing input
use skip_prefix to avoid repeating strings
use skip_prefix to avoid magic numbers
transport-helper: avoid reading past end-of-string
fast-import: fix read of uninitialized argv memory
apply: use skip_prefix instead of raw addition
refactor skip_prefix to return a boolean
avoid using skip_prefix as a boolean
daemon: mark some strings as const
parse_diff_color_slot: drop ofs parameter
In some cases, we use starts_with to check for a prefix, and
then use an already-calculated prefix length to advance a
pointer past the prefix. There are no magic numbers or
duplicated strings here, but we can still make the code
simpler and more obvious by using skip_prefix.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
extract_content_type() could not extract a charset parameter if the
parameter is not the first one and there is a whitespace and a following
semicolon just before the parameter. For example:
text/plain; format=fixed ;charset=utf-8
And it also could not handle correctly some other cases, such as:
text/plain; charset=utf-8; format=fixed
text/plain; some-param="a long value with ;semicolons;"; charset=utf-8
Thanks-to: Jeff King <peff@peff.net>
Signed-off-by: Yi EungJun <eungjun.yi@navercorp.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is specified by RFC 2616 as the default if no "charset"
parameter is given.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since the previous commit, we now give a sanitized,
shortened version of the content-type header to any callers
who ask for it.
This patch adds back a way for them to cleanly access
specific parameters to the type. We could easily extract all
parameters and make them available via a string_list, but:
1. That complicates the interface and memory management.
2. In practice, no planned callers care about anything
except the charset.
This patch therefore goes with the simplest thing, and we
can expand or change the interface later if it becomes
necessary.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we get a content-type from curl, we get the whole
header line, including any parameters, and without any
normalization (like downcasing or whitespace) applied.
If we later try to match it with strcmp() or even
strcasecmp(), we may get false negatives.
This could cause two visible behaviors:
1. We might fail to recognize a smart-http server by its
content-type.
2. We might fail to relay text/plain error messages to
users (especially if they contain a charset parameter).
This patch teaches the http code to extract and normalize
just the type/subtype portion of the string. This is
technically passing out less information to the callers, who
can no longer see the parameters. But none of the current
callers cares, and a future patch will add back an
easier-to-use method for accessing those parameters.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* mh/object-code-cleanup:
sha1_file.c: document a bunch of functions defined in the file
sha1_file_name(): declare to return a const string
find_pack_entry(): document last_found_pack
replace_object: use struct members instead of an array
Change the return value of sha1_file_name() to (const char *).
(Callers have no business mucking about here.) Change callers
accordingly, deleting a few superfluous temporary variables along the
way.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We currently don't reuse http connections when fetching via
the smart-http protocol. This is bad because the TCP
handshake introduces latency, and especially because SSL
connection setup may be non-trivial.
We can fix it by consistently using curl's "multi"
interface. The reason is rather complicated:
Our http code has two ways of being used: queuing many
"slots" to be fetched in parallel, or fetching a single
request in a blocking manner. The parallel code is built on
curl's "multi" interface. Most of the single-request code
uses http_request, which is built on top of the parallel
code (we just feed it one slot, and wait until it finishes).
However, one could also accomplish the single-request scheme
by avoiding curl's multi interface entirely and just using
curl_easy_perform. This is simpler, and is used by post_rpc
in the smart-http protocol.
It does work to use the same curl handle in both contexts,
as long as it is not at the same time. However, internally
curl may not share all of the cached resources between both
contexts. In particular, a connection formed using the
"multi" code will go into a reuse pool connected to the
"multi" object. Further requests using the "easy" interface
will not be able to reuse that connection.
The smart http protocol does ref discovery via http_request,
which uses the "multi" interface, and then follows up with
the "easy" interface for its rpc calls. As a result, we make
two HTTP connections rather than reusing a single one.
We could teach the ref discovery to use the "easy"
interface. But it is only once we have done this discovery
that we know whether the protocol will be smart or dumb. If
it is dumb, then our further requests, which want to fetch
objects in parallel, will not be able to reuse the same
connection.
Instead, this patch switches post_rpc to build on the
parallel interface, which means that we use it consistently
everywhere. It's a little more complicated to use, but since
we have the infrastructure already, it doesn't add any code;
we can just factor out the relevant bits from http_request.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Remove a few duplicate implementations of prefix/suffix comparison
functions, and rename them to starts_with and ends_with.
* cc/starts-n-ends-with:
replace {pre,suf}fixcmp() with {starts,ends}_with()
strbuf: introduce starts_with() and ends_with()
builtin/remote: remove postfixcmp() and use suffixcmp() instead
environment: normalize use of prefixcmp() by removing " != 0"
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>
Issue "100 Continue" responses to help use of GSS-Negotiate
authentication scheme over HTTP transport when needed.
* bc/http-100-continue:
remote-curl: fix large pushes with GSSAPI
remote-curl: pass curl slot_results back through run_slot
http: return curl's AUTHAVAIL via slot_results
Callers of the http code may want to know which auth types
were available for the previous request. But after finishing
with the curl slot, they are not supposed to look at the
curl handle again. We already handle returning other
information via the slot_results struct; let's add a flag to
check the available auth.
Note that older versions of curl did not support this, so we
simply return 0 (something like "-1" would be worse, as the
value is a bitflag and we might accidentally set a flag).
This is sufficient for the callers planned in this series,
who only trigger some optional behavior if particular bits
are set, and can live with a fake "no bits" answer.
Signed-off-by: Jeff King <peff@peff.net>
Handle the case where http transport gets redirected during the
authorization request better.
* jk/http-auth-redirects:
http.c: Spell the null pointer as NULL
remote-curl: rewrite base url from info/refs redirects
remote-curl: store url as a strbuf
remote-curl: make refs_url a strbuf
http: update base URLs when we see redirects
http: provide effective url to callers
http: hoist credential request out of handle_curl_result
http: refactor options to http_get_*
http_request: factor out curlinfo_strbuf
http_get_file: style fixes
Commit 1bbcc224 ("http: refactor options to http_get_*", 28-09-2013)
changed the type of final 'options' argument of the http_get_file()
function from an int to an 'struct http_get_options' pointer.
However, it neglected to update the (single) call site. Since this
call was passing '0' to that argument, it was (correctly) being
interpreted as a null pointer. Change to argument to NULL.
Noticed by sparse. ("Using plain integer as NULL pointer")
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Acked-by: Jeff King <peff@peff.net>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit a15d069 taught git to use curl's SOCKOPTFUNCTION hook
to turn on TCP keepalives. However, modern versions of curl
have a TCP_KEEPALIVE option, which can do this for us. As an
added bonus, the curl code knows how to turn on keepalive
for a much wider variety of platforms. The only downside to
using this option is that not everybody has a new enough curl.
Let's split our keepalive options into three conditionals:
1. With curl 7.25.0 and newer, we rely on curl to do it
right.
2. With older curl that still knows SOCKOPTFUNCTION, we
use the code from a15d069.
3. Otherwise, we are out of luck, and the call is a no-op.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If a caller asks the http_get_* functions to go to a
particular URL and we end up elsewhere due to a redirect,
the effective_url field can tell us where we went.
It would be nice to remember this redirect and short-cut
further requests for two reasons:
1. It's more efficient. Otherwise we spend an extra http
round-trip to the server for each subsequent request,
just to get redirected.
2. If we end up with an http 401 and are going to ask for
credentials, it is to feed them to the redirect target.
If the redirect is an http->https upgrade, this means
our credentials may be provided on the http leg, just
to end up redirected to https. And if the redirect
crosses server boundaries, then curl will drop the
credentials entirely as it follows the redirect.
However, it, it is not enough to simply record the effective
URL we saw and use that for subsequent requests. We were
originally fed a "base" url like:
http://example.com/foo.git
and we want to figure out what the new base is, even though
the URLs we see may be:
original: http://example.com/foo.git/info/refs
effective: http://example.com/bar.git/info/refs
Subsequent requests will not be for "info/refs", but for
other paths relative to the base. We must ask the caller to
pass in the original base, and we must pass the redirected
base back to the caller (so that it can generate more URLs
from it). Furthermore, we need to feed the new base to the
credential code, so that requests to credential helpers (or
to the user) match the URL we will be requesting.
This patch teaches http_request_reauth to do this munging.
Since it is the caller who cares about making more URLs, it
seems at first glance that callers could simply check
effective_url themselves and handle it. However, since we
need to update the credential struct before the second
re-auth request, we have to do it inside http_request_reauth.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When we ask curl to access a URL, it may follow one or more
redirects to reach the final location. We have no idea
this has happened, as curl takes care of the details and
simply returns the final content to us.
The final URL that we ended up with can be accessed via
CURLINFO_EFFECTIVE_URL. Let's make that optionally available
to callers of http_get_*, so that they can make further
decisions based on the redirection.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When we are handling a curl response code in http_request or
in the remote-curl RPC code, we use the handle_curl_result
helper to translate curl's response into an easy-to-use
code. When we see an HTTP 401, we do one of two things:
1. If we already had a filled-in credential, we mark it as
rejected, and then return HTTP_NOAUTH to indicate to
the caller that we failed.
2. If we didn't, then we ask for a new credential and tell
the caller HTTP_REAUTH to indicate that they may want
to try again.
Rejecting in the first case makes sense; it is the natural
result of the request we just made. However, prompting for
more credentials in the second step does not always make
sense. We do not know for sure that the caller is going to
make a second request, and nor are we sure that it will be
to the same URL. Logically, the prompt belongs not to the
request we just finished, but to the request we are (maybe)
about to make.
In practice, it is very hard to trigger any bad behavior.
Currently, if we make a second request, it will always be to
the same URL (even in the face of redirects, because curl
handles the redirects internally). And we almost always
retry on HTTP_REAUTH these days. The one exception is if we
are streaming a large RPC request to the server (e.g., a
pushed packfile), in which case we cannot restart. It's
extremely unlikely to see a 401 response at this stage,
though, as we would typically have seen it when we sent a
probe request, before streaming the data.
This patch drops the automatic prompt out of case 2, and
instead requires the caller to do it. This is a few extra
lines of code, and the bug it fixes is unlikely to come up
in practice. But it is conceptually cleaner, and paves the
way for better handling of credentials across redirects.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
This is a follow up to commit e47a8583 (enable SO_KEEPALIVE for
connected TCP sockets, 2011-12-06).
Sockets may never receive notification of some link errors,
causing "git fetch" or similar processes to hang forever.
Enabling keepalive messages allows hung processes to error out
after a few minutes/hours depending on the keepalive settings of
the system.
I noticed this problem with some non-interactive cronjobs getting
hung when talking to HTTP servers.
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Over time, the http_get_strbuf function has grown several
optional parameters. We now have a bitfield with multiple
boolean options, as well as an optional strbuf for returning
the content-type of the response. And a future patch in this
series is going to add another strbuf option.
Treating these as separate arguments has a few downsides:
1. Most call sites need to add extra NULLs and 0s for the
options they aren't interested in.
2. The http_get_* functions are actually wrappers around
2 layers of low-level implementation functions. We have
to pass these options through individually.
3. The http_get_strbuf wrapper learned these options, but
nobody bothered to do so for http_get_file, even though
it is backed by the same function that does understand
the options.
Let's consolidate the options into a single struct. For the
common case of the default options, we'll allow callers to
simply pass a NULL for the options struct.
The resulting code is often a few lines longer, but it ends
up being easier to read (and to change as we add new
options, since we do not need to update each call site).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When we retrieve the content-type of an http response, curl
gives us a pointer to internal storage, which we then copy
into a strbuf. Let's factor out the get-and-copy routine,
which can be used for getting other curl info.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Besides being ugly, the extra parentheses are idiomatic for
suppressing compiler warnings when we are assigning within a
conditional. We aren't doing that here, and they just
confuse the reader.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Allow section.<urlpattern>.var configuration variables to be
treated as a "virtual" section.var given a URL, and use the
mechanism to enhance http.* configuration variables.
This is a reroll of Kyle J. McKay's work.
* jc/url-match:
builtin/config.c: compilation fix
config: "git config --get-urlmatch" parses section.<url>.key
builtin/config: refactor collect_config()
config: parse http.<url>.<variable> using urlmatch
config: add generic callback wrapper to parse section.<url>.key
config: add helper to normalize and match URLs
http.c: fix parsing of http.sslCertPasswordProtected variable
Use the urlmatch_config_entry() to wrap the underlying
http_options() two-level variable parser in order to set
http.<variable> to the value with the most specific URL in the
configuration.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Kyle J. McKay <mackyle@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The existing code triggers only when the configuration variable is
set to true. Once the variable is set to true in a more generic
configuration file (e.g. ~/.gitconfig), it cannot be overriden to
false in the repository specific one (e.g. .git/config).
Signed-off-by: Junio C Hamano <gitster@pobox.com>
HTTP servers may send Set-Cookie headers in a response and expect them
to be set on subsequent requests. By default, libcurl behavior is to
store such cookies in memory and reuse them across requests within a
single session. However, it may also make sense, depending on the
server and the cookies, to store them across sessions. Provide users
an option to enable this behavior, writing cookies out to the same
file specified in http.cookiefile.
Signed-off-by: Dave Borowitz <dborowitz@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Older cURL wanted piece of memory we call it with to be stable, but
we updated the auth material after handing it to a call.
* bc/http-keep-memory-given-to-curl:
http.c: don't rewrite the user:passwd string multiple times
Curl older than 7.17 (RHEL 4.X provides 7.12 and RHEL 5.X provides
7.15) requires that we manage any strings that we pass to it as
pointers. So, we really shouldn't be modifying this strbuf after we
have passed it to curl.
Our interaction with curl is currently safe (before or after this
patch) since the pointer that is passed to curl is never invalidated;
it is repeatedly rewritten with the same sequence of characters but
the strbuf functions never need to allocate a larger string, so the
same memory buffer is reused.
This "guarantee" of safety is somewhat subtle and could be overlooked
by someone who may want to add a more complex handling of the username
and password. So, let's stop modifying this strbuf after we have
passed it to curl, but also leave a note to describe the assumptions
that have been made about username/password lifetime and to draw
attention to the code.
Signed-off-by: Brandon Casey <drafnel@gmail.com>
Acked-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Because we reuse curl handles for multiple requests, the
setup of a handle happens in two stages: stable, global
setup and per-request setup. The lifecycle of a handle is
something like:
1. get_curl_handle; do basic global setup that will last
through the whole program (e.g., setting the user
agent, ssl options, etc)
2. get_active_slot; set up a per-request baseline (e.g.,
clearing the read/write functions, making it a GET
request, etc)
3. perform the request with curl_*_perform functions
4. goto step 2 to perform another request
Breaking it down this way means we can avoid doing global
setup from step (1) repeatedly, but we still finish step (2)
with a predictable baseline setup that callers can rely on.
Until commit 6d052d7 (http: add HTTP_KEEP_ERROR option,
2013-04-05), setting curl's FAILONERROR option was a global
setup; we never changed it. However, 6d052d7 introduced an
option where some requests might turn off FAILONERROR. Later
requests using the same handle would have the option
unexpectedly turned off, which meant they would not notice
http failures at all.
This could easily be seen in the test-suite for the
"half-auth" cases of t5541 and t5551. The initial requests
turned off FAILONERROR, which meant it was erroneously off
for the rpc POST. That worked fine for a successful request,
but meant that we failed to react properly to the HTTP 401
(instead, we treated whatever the server handed us as a
successful message body).
The solution is simple: now that FAILONERROR is a
per-request setting, we move it to get_active_slot to make
sure it is reset for each request.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a boolean http.sslTry option which allows to enable AUTH SSL/TLS and
encrypted data transfers when connecting via regular FTP protocol.
Default is false since it might trigger certificate verification errors on
misconfigured servers.
Signed-off-by: Modestas Vainius <modestas@vainius.eu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This function is a single-liner and is only called from one
place. Just inline it, which makes the code more obvious.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we report an http error code, we say something like:
error: The requested URL reported failure: 403 Forbidden while accessing http://example.com/repo.git
Everything between "error:" and "while" is written by curl,
and the resulting sentence is hard to read (especially
because there is no punctuation between curl's sentence and
the remainder of ours). Instead, let's re-order this to give
better flow:
error: unable to access 'http://example.com/repo.git: The requested URL reported failure: 403 Forbidden
This is still annoyingly long, but at least reads more
clearly left to right.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This helper function should really be a one-liner that
prints an error message, but it has ended up unnecessarily
complicated:
1. We call error() directly when we fail to start the curl
request, so we must later avoid printing a duplicate
error in http_error().
It would be much simpler in this case to just stuff the
error message into our usual curl_errorstr buffer
rather than printing it ourselves. This means that
http_error does not even have to care about curl's exit
value (the interesting part is in the errorstr buffer
already).
2. We return the "ret" value passed in to us, but none of
the callers actually cares about our return value. We
can just drop this entirely.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We currently set curl's FAILONERROR option, which means that
any http failures are reported as curl errors, and the
http body content from the server is thrown away.
This patch introduces a new option to http_get_strbuf which
specifies that the body content from a failed http response
should be placed in the destination strbuf, where it can be
accessed by the caller.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Having the packet sizes defined near the packet read/write
functions makes more sense.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Callers may pass us a strbuf which we use to record the
content-type of the response. However, we simply appended to
it rather than overwriting its contents, meaning that cruft
in the strbuf gave us a bogus type. E.g., the multiple
requests triggered by http_request could yield a type like
"text/plainapplication/x-git-receive-pack-advertisement".
Reported-by: Michael Schubert <mschub@elegosoft.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Before parsing a suspected smart-HTTP response verify the returned
Content-Type matches the standard. This protects a client from
attempting to process a payload that smells like a smart-HTTP
server response.
JGit has been doing this check on all responses since the dawn of
time. I mistakenly failed to include it in git-core when smart HTTP
was introduced. At the time I didn't know how to get the Content-Type
from libcurl. I punted, meant to circle back and fix this, and just
plain forgot about it.
Signed-off-by: Shawn Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If sslCertPasswordProtected is set to true do not ask for username to decrypt rsa key. This question is pointless, the key is only protected by a password. Internaly the username is simply set to "".
Signed-off-by: Rene Bredlau <git@unrelated.de>
Acked-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Sometimes curl_multi_timeout() function suggested a wrong timeout
value when there is no file descriptors to wait on and the http
transport ended up sleeping for minutes in select(2) system call.
Detect this and reduce the wait timeout in such a case.
* sz/maint-curl-multi-timeout:
Fix potential hang in https handshake
Further clean-up to the http codepath that picks up results after
cURL library is done with one request slot.
* jk/maint-http-init-not-in-result-handler:
http: do not set up curl auth after a 401
remote-curl: do not call run_slot repeatedly
It has been observed that curl_multi_timeout may return a very long
timeout value (e.g., 294 seconds and some usec) just before
curl_multi_fdset returns no file descriptors for reading. The
upshot is that select() will hang for a long time -- long enough for
an https handshake to be dropped. The observed behavior is that
the git command will hang at the terminal and never transfer any
data.
This patch is a workaround for a probable bug in libcurl. The bug
only seems to manifest around a very specific set of circumstances:
- curl version (from curl/curlver.h):
#define LIBCURL_VERSION_NUM 0x071307
- git-remote-https running on an ubuntu-lucid VM.
- Connecting through squid proxy running on another VM.
Interestingly, the problem doesn't manifest if a host connects
through squid proxy running on localhost; only if the proxy is on
a separate VM (not sure if the squid host needs to be on a separate
physical machine). That would seem to suggest that this issue
is timing-sensitive.
This patch is more or less in line with a recommendation in the
curl docs about how to behave when curl_multi_fdset doesn't return
and file descriptors:
http://curl.haxx.se/libcurl/c/curl_multi_fdset.html
Signed-off-by: Stefan Zager <szager@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Fixes a regression in maint-1.7.11 (v1.7.11.7), maint (v1.7.12.1)
and master (v1.8.0-rc0).
* jk/maint-http-half-auth-push:
http: fix segfault in handle_curl_result
When we get an http 401, we prompt for credentials and put
them in our global credential struct. We also feed them to
the curl handle that produced the 401, with the intent that
they will be used on a retry.
When the code was originally introduced in commit 42653c0,
this was a necessary step. However, since dfa1725, we always
feed our global credential into every curl handle when we
initialize the slot with get_active_slot. So every further
request already feeds the credential to curl.
Moreover, accessing the slot here is somewhat dubious. After
the slot has produced a response, we don't actually control
it any more. If we are using curl_multi, it may even have
been re-initialized to handle a different request.
It just so happens that we will reuse the curl handle within
the slot in such a case, and that because we only keep one
global credential, it will be the one we want. So the
current code is not buggy, but it is misleading.
By cleaning it up, we can remove the slot argument entirely
from handle_curl_result, making it much more obvious that
slots should not be accessed after they are marked as
finished.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we create an http active_request_slot, we can set its
"results" pointer back to local storage. The http code will
fill in the details of how the request went, and we can
access those details even after the slot has been cleaned
up.
Commit 8809703 (http: factor out http error code handling)
switched us from accessing our local results struct directly
to accessing it via the "results" pointer of the slot. That
means we're accessing the slot after it has been marked as
finished, defeating the whole purpose of keeping the results
storage separate.
Most of the time this doesn't matter, as finishing the slot
does not actually clean up the pointer. However, when using
curl's multi interface with the dumb-http revision walker,
we might actually start a new request before handing control
back to the original caller. In that case, we may reuse the
slot, zeroing its results pointer, and leading the original
caller to segfault while looking for its results inside the
slot.
Instead, we need to pass a pointer to our local results
storage to the handle_curl_result function, rather than
relying on the pointer in the slot struct. This matches what
the original code did before the refactoring (which did not
use a separate function, and therefore just accessed the
results struct directly).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some HTTP servers try to use gzip compression on the /info/refs
request to save transfer bandwidth. Repositories with many tags
may find the /info/refs request can be gzipped to be 50% of the
original size due to the few but often repeated bytes used (hex
SHA-1 and commonly digits in tag names).
For most HTTP requests enable "Accept-Encoding: gzip" ensuring
the /info/refs payload can use this encoding format.
Only request gzip encoding from servers. Although deflate is
supported by libcurl, most servers have standardized on gzip
encoding for compression as that is what most browsers support.
Asking for deflate increases request sizes by a few bytes, but is
unlikely to ever be used by a server.
Disable the Accept-Encoding header on probe RPCs as response bodies
are supposed to be exactly 4 bytes long, "0000". The HTTP headers
requesting and indicating compression use more space than the data
transferred in the body.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
