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>
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>
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>
Most of our http requests go through the http_request()
interface, which does some nice post-processing on the
results. In particular, it handles prompting for missing
credentials as well as approving and rejecting valid or
invalid credentials. Unfortunately, it only handles GET
requests. Making it handle POSTs would be quite complex, so
let's pull result handling code into its own function so
that it can be reused from the POST code paths.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Before commit 986bbc08, git was proactive about asking for
http passwords. It assumed that if you had a username in
your URL, you would also want a password, and asked for it
before making any http requests.
However, this could interfere with the use of .netrc (see
986bbc08 for details). And it was also unnecessary, since
the http fetching code had learned to recognize an HTTP 401
and prompt the user then. Furthermore, the proactive prompt
could interfere with the usage of .netrc (see 986bbc08 for
details).
Unfortunately, the http push-over-DAV code never learned to
recognize HTTP 401, and so was broken by this change. This
patch does a quick fix of re-enabling the "proactive auth"
strategy only for http-push, leaving the dumb http fetch and
smart-http as-is.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* mf/curl-select-fdset:
http: drop "local" member from request struct
http.c: Rely on select instead of tracking whether data was received
http.c: Use timeout suggested by curl instead of fixed 50ms timeout
http.c: Use curl_multi_fdset to select on curl fds instead of just sleeping
This is a FILE pointer in the case that we are sending our
output to a file. We originally used it to run ftell() to
determine whether data had been written to our file during
our last call to curl. However, as of the last patch, we no
longer care about that flag anymore. All uses of this struct
member are now just book-keeping that can go away.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since now select is used with the file descriptors of the http connections,
tracking whether data was received recently (and trying to read more in
that case) is no longer necessary. Instead, always call select and rely on
it to return as soon as new data can be read.
Signed-off-by: Mika Fischer <mika.fischer@zoopnet.de>
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The http_init function takes a "struct remote". Part of its
initialization procedure is to look at the remote's url and
grab some auth-related parameters. However, using the url
included in the remote is:
- wrong; the remote-curl helper may have a separate,
unrelated URL (e.g., from remote.*.pushurl). Looking at
the remote's configured url is incorrect.
- incomplete; http-fetch doesn't have a remote, so passes
NULL. So http_init never gets to see the URL we are
actually going to use.
- cumbersome; http-push has a similar problem to
http-fetch, but actually builds a fake remote just to
pass in the URL.
Instead, let's just add a separate URL parameter to
http_init, and all three callsites can pass in the
appropriate information.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jc/zlib-wrap:
zlib: allow feeding more than 4GB in one go
zlib: zlib can only process 4GB at a time
zlib: wrap deflateBound() too
zlib: wrap deflate side of the API
zlib: wrap inflateInit2 used to accept only for gzip format
zlib: wrap remaining calls to direct inflate/inflateEnd
zlib wrapper: refactor error message formatter
Conflicts:
sha1_file.c
The size of objects we read from the repository and data we try to put
into the repository are represented in "unsigned long", so that on larger
architectures we can handle objects that weigh more than 4GB.
But the interface defined in zlib.h to communicate with inflate/deflate
limits avail_in (how many bytes of input are we calling zlib with) and
avail_out (how many bytes of output from zlib are we ready to accept)
fields effectively to 4GB by defining their type to be uInt.
In many places in our code, we allocate a large buffer (e.g. mmap'ing a
large loose object file) and tell zlib its size by assigning the size to
avail_in field of the stream, but that will truncate the high octets of
the real size. The worst part of this story is that we often pass around
z_stream (the state object used by zlib) to keep track of the number of
used bytes in input/output buffer by inspecting these two fields, which
practically limits our callchain to the same 4GB limit.
Wrap z_stream in another structure git_zstream that can express avail_in
and avail_out in unsigned long. For now, just die() when the caller gives
a size that cannot be given to a single zlib call. In later patches in the
series, we would make git_inflate() and git_deflate() internally loop to
give callers an illusion that our "improved" version of zlib interface can
operate on a buffer larger than 4GB in one go.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Yes, these don't match perfectly with the void* first parameter of the
fread/fwrite in the standard library, but they do match the curl
expected method signature. This is needed when a refactor passes a
curl_write_callback around, which would otherwise give incorrect
parameter warnings.
Signed-off-by: Dan McGee <dpmcgee@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In a struct definitions, unlike functions, the prevailing style is for
the opening brace to go on the same line as the struct name, like so:
struct foo {
int bar;
char *baz;
};
Indeed, grepping for 'struct [a-z_]* {$' yields about 5 times as many
matches as 'struct [a-z_]*$'.
Linus sayeth:
Heretic people all over the world have claimed that this inconsistency
is ... well ... inconsistent, but all right-thinking people know that
(a) K&R are _right_ and (b) K&R are right.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This allows non-http/curl users to access it too (eg. http-backend.c).
Update include headers in end_url_with_slash() users too.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* sp/maint-dumb-http-pack-reidx:
http.c::new_http_pack_request: do away with the temp variable filename
http-fetch: Use temporary files for pack-*.idx until verified
http-fetch: Use index-pack rather than verify-pack to check packs
Allow parse_pack_index on temporary files
Extract verify_pack_index for reuse from verify_pack
Introduce close_pack_index to permit replacement
http.c: Remove unnecessary strdup of sha1_to_hex result
http.c: Don't store destination name in request structures
http.c: Drop useless != NULL test in finish_http_pack_request
http.c: Tiny refactoring of finish_http_pack_request
t5550-http-fetch: Use subshell for repository operations
http.c: Remove bad free of static block
* rc/maint-curl-helper:
remote-curl: ensure that URLs have a trailing slash
http: make end_url_with_slash() public
t5541-http-push: add test for URLs with trailing slash
Conflicts:
remote-curl.c
The destination name within the object store is easily computed
on demand, reusing a static buffer held by sha1_file.c. We don't
need to copy the entire path into the request structure for safe
keeping, when it can be easily reformatted after the download has
been completed.
This reduces the size of the per-request structure, and removes
yet another PATH_MAX based limit.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When an HTTP request returns a 401, Git will currently fail with a
confusing message saying that it got a 401, which is not very
descriptive.
Currently if a user wants to use Git over HTTP, they have to use one
URL with the username in the URL (e.g. "http://user@host.com/repo.git")
for write access and another without the username for unauthenticated
read access (unless they want to be prompted for the password each
time). However, since the HTTP servers will return a 401 if an action
requires authentication, we can prompt for username and password if we
see this, allowing us to use a single URL for both purposes.
This patch changes http_request to prompt for the username and password,
then return HTTP_REAUTH so http_get_strbuf can try again. If it gets
a 401 even when a user/pass is supplied, http_request will now return
HTTP_NOAUTH which remote_curl can then use to display a more
intelligent error message that is less confusing.
Signed-off-by: Scott Chacon <schacon@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The git-remote-curl backend detects if the remote server supports
the git-receive-pack service, and if so, runs git-send-pack in a
pipe to dump the command and pack data as a single POST request.
The advertisements from the server that were obtained during the
discovery are passed into git-send-pack before the POST request
starts. This permits git-send-pack to operate largely unmodified.
For smaller packs (those under 1 MiB) a HTTP/1.0 POST with a
Content-Length is used, permitting interaction with any server.
The 1 MiB limit is arbitrary, but is sufficent to fit most deltas
created by human authors against text sources with the occasional
small binary file (e.g. few KiB icon image). The configuration
option http.postBuffer can be used to increase (or shink) this
buffer if the default is not sufficient.
For larger packs which cannot be spooled entirely into the helper's
memory space (due to http.postBuffer being too small), the POST
request requires HTTP/1.1 and sets "Transfer-Encoding: chunked".
This permits the client to upload an unknown amount of data in one
HTTP transaction without needing to pregenerate the entire pack
file locally.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
CC: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The code handling the fetching of loose objects in http-push.c and
http-walker.c have been refactored into new methods and a new struct
(object_http_request) in http.c. They are not meant to be invoked
elsewhere.
The new methods in http.c are
- new_http_object_request
- process_http_object_request
- finish_http_object_request
- abort_http_object_request
- release_http_object_request
and the new struct is http_object_request.
RANGER_HEADER_SIZE and no_pragma_header is no longer made available
outside of http.c, since after the above changes, there are no other
instances of usage outside of http.c.
Remove members of the transfer_request struct in http-push.c and
http-walker.c, including filename, real_sha1 and zret, as they are used
no longer used.
Move the methods append_remote_object_url() and get_remote_object_url()
from http-push.c to http.c. Additionally, get_remote_object_url() is no
longer defined only when USE_CURL_MULTI is defined, since
non-USE_CURL_MULTI code in http.c uses it (namely, in
new_http_object_request()).
Refactor code from http-push.c::start_fetch_loose() and
http-walker.c::start_object_fetch_request() that deals with the details
of coming up with the filename to store the retrieved object, resuming
a previously aborted request, and making a new curl request, into a new
function, new_http_object_request().
Refactor code from http-walker.c::process_object_request() into the
function, process_http_object_request().
Refactor code from http-push.c::finish_request() and
http-walker.c::finish_object_request() into a new function,
finish_http_object_request(). It returns the result of the
move_temp_to_file() invocation.
Add a function, release_http_object_request(), which cleans up object
request data. http-push.c and http-walker.c invoke this function
separately; http-push.c::release_request() and
http-walker.c::release_object_request() do not invoke this function.
Add a function, abort_http_object_request(), which unlink()s the object
file and invokes release_http_object_request(). Update
http-walker.c::abort_object_request() to use this.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The code handling the fetching of packs in http-push.c and
http-walker.c have been refactored into new methods and a new struct
(http_pack_request) in http.c. They are not meant to be invoked
elsewhere.
The new methods in http.c are
- new_http_pack_request
- finish_http_pack_request
- release_http_pack_request
and the new struct is http_pack_request.
Add a function, new_http_pack_request(), that deals with the details of
coming up with the filename to store the retrieved packfile, resuming a
previously aborted request, and making a new curl request. Update
http-push.c::start_fetch_packed() and http-walker.c::fetch_pack() to
use this.
Add a function, finish_http_pack_request(), that deals with renaming
the pack, advancing the pack list, and installing the pack. Update
http-push.c::finish_request() and http-walker.c::fetch_pack to use
this.
Update release_request() in http-push.c and http-walker.c to invoke
release_http_pack_request() to clean up pack request helper data.
The local_stream member of the transfer_request struct in http-push.c
has been removed, as the packfile pointer will be managed in the struct
http_pack_request.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
http-push.c and http-walker.c no longer have to use fetch_index or
setup_index; they simply need to use http_get_info_packs, a new http
method, in their fetch_indices implementations.
Move fetch_index() and rename to fetch_pack_index() in http.c; this
method is not meant to be used outside of http.c. It invokes
end_url_with_slash with base_url; apart from that change, the code is
identical.
Move setup_index() and rename to fetch_and_setup_pack_index() in
http.c; this method is not meant to be used outside of http.c.
Do not immediately set ret to 0 in http-walker.c::fetch_indices();
instead do it in the HTTP_MISSING_TARGET case, to make it clear that
the HTTP_OK and HTTP_MISSING_TARGET cases both return 0.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The new functions added are:
- http_request() (internal function)
- http_get_strbuf()
- http_get_file()
- http_error()
http_get_strbuf and http_get_file allow respectively to retrieve contents of
an URL to a strbuf or an opened file handle.
http_error prints out an error message containing the URL and the curl error
(in curl_errorstr).
Signed-off-by: Mike Hommey <mh@glandium.org>
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move RANGE_HEADER_SIZE to http.h.
Create no_pragma_header, the curl header list containing the header
"Pragma:" in http.[ch]. It is allocated in http_init, and freed in
http_cleanup. This replaces the no_pragma_header in http-push.c, and
the no_pragma_header member in walker_data in http-walker.c.
Create http_is_verbose. It is to be used by methods in http.c, and is
modified at the entry points of http.c's users, namely http-push.c
(when parsing options) and http-walker.c (in get_http_walker).
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When using multi-pass authentication methods, the curl library may
need to rewind the read buffers (depending on how much already has
been fed to the server) used for providing data to HTTP PUT, POST or
PROPFIND, and in order to allow the library to do so, we need to tell
it how by providing either an ioctl callback or a seek callback.
This patch adds an ioctl callback, which should be usable on older
curl versions (since 7.12.3) than the seek callback (introduced in
curl 7.18.0).
Some HTTP servers (such as Apache) give an 401 error reply immediately
after receiving the headers (so no data has been read from the read
buffers, and thus no rewinding is needed), but other servers (such
as Lighttpd) only replies after the whole request has been sent and
all data has been read from the read buffers, making rewinding necessary.
Signed-off-by: Martin Storsjo <martin@martin.st>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
After master.k.org upgrade, I started seeing these warning messages:
transport.c: In function 'get_refs_via_curl':
transport.c:458: error: call to '_curl_easy_setopt_err_write_callback' declared with attribute warning: curl_easy_setopt expects a curl_write_callback argument for this option
It appears that the curl header wants to enforce the function signature
for callback function given to curl_easy_setopt() to be compatible with
that of (*curl_write_callback) or fwrite. This patch seems to work the
issue around.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This simplifies a few things, makes a few things slightly more
complicated, but, more importantly, allows that, when struct ref can
represent a symref, http_fetch_ref() can return one.
Incidentally makes the string that http_fetch_ref() gets include "refs/"
(if appropriate), because that's how the name field of struct ref works.
As far as I can tell, the usage in walker:interpret_target() wouldn't have
worked previously, if it ever would have been used, which it wouldn't
(since the fetch process uses the hash instead of the name of the ref
there).
Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In transport.c, proxy setting (the one from the remote conf) was set through
curl_easy_setopt() call, while http.c already does the same with the
http.proxy setting. We now just use this infrastructure instead, and make
http_init() now take the struct remote as argument so that it can take the
http_proxy setting from there, and any other property that would be added
later.
At the same time, we make get_http_walker() take a struct remote argument
too, and pass it to http_init(), which makes remote defined proxy be used
for more than get_refs_via_curl().
We leave out http-fetch and http-push, which don't use remotes for the
moment, purposefully.
Signed-off-by: Mike Hommey <mh@glandium.org>
Acked-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Make the necessary changes to be ok with their difference, and rename the
function http_fetch_ref.
Signed-off-by: Mike Hommey <mh@glandium.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Quite some variables defined as extern in http.h are only used in http.c,
and some others, only defined in http.c, were not static.
Signed-off-by: Mike Hommey <mh@glandium.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This eliminates the last function provided by the code using http.h as
a global symbol, so it should be possible to have multiple programs
using http.h in the same executable, and it also adds an argument to
that callback, so that info can be passed into the callback without
being global.
Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This removes all of the boilerplate and http-internal stuff from
fill_active_slots() and makes it easy to turn into a callback.
Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
curl_multi_remove_handle() is broken in libcurl < 7.16, in that it
doesn't correctly update the active handles count when a request is
aborted. This causes the transfer to hang forever waiting for the
handle count to become less than the number of active requests.
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
It appears that curl_easy_duphandle() from libcurl 7.16.0
returns a curl session handle which fails GOOD_MULTI_HANDLE()
check in curl_multi_add_handle(). This causes fetch_ref() to
fail because start_active_slot() cannot start the request.
For now, check for 7.16.0 to work this issue around.
Signed-off-by: Junio C Hamano <junkio@cox.net>
Older curl releases do not define CURLE_HTTP_RETURNED_ERROR, they
use CURLE_HTTP_NOT_FOUND instead. Newer curl releases keep the
CURLE_HTTP_NOT_FOUND definition but using a -DCURL_NO_OLDIES
preprocessor flag the old name will not be present in the 'curl.h'
header.
This patch makes our code written for newer releases of the curl
library but allow compiling against an older curl (older than
0x070a03) by defining the missing CURLE_HTTP_RETURNED_ERROR as a
synonym for CURLE_HTTP_NOT_FOUND.
Signed-off-by: Art Haas <ahaas@airmail.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Incorporate into http-push a fix related to accessing slot results after
the slot was reused, and fix a case in run_active_slot where a
finished slot wasn't detected if the slot was reused.
Signed-off-by: Junio C Hamano <junkio@cox.net>
In fetch_object, there's a call to release an object request if the
object mysteriously arrived, say in a pack. Unfortunately, the fetch
attempt for this object might already be in progress, and we'll leak the
descriptor. Instead, try to tidy away the request.
Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Add a way to store the results of an HTTP request when a slot finishes
so the results can be processed after the slot has been reused.
Signed-off-by: Nick Hengeveld <nickh@reactrix.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
Move shared HTTP request functionality out of http-fetch and http-push,
and replace the two fwrite_buffer/fwrite_buffer_dynamic functions with
one fwrite_buffer function that does dynamic buffering. Use slot
callbacks to process responses to fetch object transfer requests and
push transfer requests, and put all of http-push into an #ifdef check
for curl multi support.
Signed-off-by: Nick Hengeveld <nickh@reactrix.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
