These verfy that the 'memory tracking' subsystem is actually doing its
job when using curl tool (#96), a test in libtest (#558) and also a unit
test (#1330), in order to prevent regressions in this functionallity.
Introducing a number of options to the multi interface that
allows for multiple pipelines to the same host, in order to
optimize the balance between the penalty for opening new
connections and the potential pipelining latency.
Two new options for limiting the number of connections:
CURLMOPT_MAX_HOST_CONNECTIONS - Limits the number of running connections
to the same host. When adding a handle that exceeds this limit,
that handle will be put in a pending state until another handle is
finished, so we can reuse the connection.
CURLMOPT_MAX_TOTAL_CONNECTIONS - Limits the number of connections in total.
When adding a handle that exceeds this limit,
that handle will be put in a pending state until another handle is
finished. The free connection will then be reused, if possible, or
closed if the pending handle can't reuse it.
Several new options for pipelining:
CURLMOPT_MAX_PIPELINE_LENGTH - Limits the pipeling length. If a
pipeline is "full" when a connection is to be reused, a new connection
will be opened if the CURLMOPT_MAX_xxx_CONNECTIONS limits allow it.
If not, the handle will be put in a pending state until a connection is
ready (either free or a pipe got shorter).
CURLMOPT_CONTENT_LENGTH_PENALTY_SIZE - A pipelined connection will not
be reused if it is currently processing a transfer with a content
length that is larger than this.
CURLMOPT_CHUNK_LENGTH_PENALTY_SIZE - A pipelined connection will not
be reused if it is currently processing a chunk larger than this.
CURLMOPT_PIPELINING_SITE_BL - A blacklist of hosts that don't allow
pipelining.
CURLMOPT_PIPELINING_SERVER_BL - A blacklist of server types that don't allow
pipelining.
See the curl_multi_setopt() man page for details.
This bug report properly identified that when doing SMTP and aborting
the transfer with a callback, it must be considered aborted prematurely
by the code to avoid QUIT etc to be attempted as that would cause a
hang.
The new test case 1507 verifies this behavior.
Reported by: Patricia Muscalu
Bug: http://curl.haxx.se/bug/view.cgi?id=1184
When a connection is no longer used, it is kept in the cache. If the
cache is full, the oldest idle connection is closed. If no connection is
idle, the current one is closed instead.
Automake documents that doing this will make it choose a different name
for intermediate object files even when sharing source files across
targets of same Makefile.am.
Up to automake 1.13.1 target's intermediate object files were placed
in the build subdirectory of the target. We depended on this, probably
undocumented behavior, to achieve same behavior as if a per-target flag
had been specified when building targets that actually belong to
different Makefile.am files.
It seems automake 1.13.2 is going to break behavior mentioned above.
So, lets use a documented behavior in order to achieve same purpose,
across automake versions, no matter where automake wishes to place
intermediate object files.
Our build targets that already were using a per-target '_CFLAGS' or
'_CPPFLAGS' need no 'fixing', these were already 'fixed'. The only
Makefile.am or Makefile.in files in libcurl's source tree touched by
this 'fix' are tests/libtest/Makefile.inc and tests/unit/Makefile.inc.
This reverts renaming and usage of lib/*.h header files done
28-12-2012, reverting 2 commits:
f871de0... build: make use of 76 lib/*.h renamed files
ffd8e12... build: rename 76 lib/*.h files
This also reverts removal of redundant include guard (redundant thanks
to changes in above commits) done 2-12-2013, reverting 1 commit:
c087374... curl_setup.h: remove redundant include guard
This also reverts renaming and usage of lib/*.c source files done
3-12-2013, reverting 3 commits:
13606bb... build: make use of 93 lib/*.c renamed files
5b6e792... build: rename 93 lib/*.c files
7d83dff... build: commit 13606bbfde follow-up 1
Start of related discussion thread:
http://curl.haxx.se/mail/lib-2013-01/0012.html
Asking for confirmation on pushing this revertion commit:
http://curl.haxx.se/mail/lib-2013-01/0048.html
Confirmation summary:
http://curl.haxx.se/mail/lib-2013-01/0079.html
NOTICE: The list of 2 files that have been modified by other
intermixed commits, while renamed, and also by at least one
of the 6 commits this one reverts follows below. These 2 files
will exhibit a hole in history unless git's '--follow' option
is used when viewing logs.
lib/curl_imap.h
lib/curl_smtp.h
DNS cache entries populated with CURLOPT_RESOLVE were not properly freed
again when done using the multi interface.
Test case 1502 added to verify.
Bug: http://curl.haxx.se/bug/view.cgi?id=3575448
Reported by: Alex Gruz
As pointed out in Bug report #3579064, curl_multi_perform() would
wrongly use a blocking mechanism internally for some commands which
could lead to for example a very long block if the LIST response never
showed.
The solution was to make sure to properly continue to use the multi
interface non-blocking state machine.
The new test 1501 verifies the fix.
Bug: http://curl.haxx.se/bug/view.cgi?id=3579064
Reported by: Guido Berhoerster
With FOLLOWLOCATION enabled. When a 3xx page is downloaded and the
download size was known (like with a Content-Length header), but the
subsequent URL (transfered after the 3xx page) was chunked encoded, then
the previous "known download size" would linger and cause the progress
meter to get incorrect information, ie the former value would remain
being sent in. This could easily result in downloads that were WAY
larger than "expected" and would cause >100% outputs with the curl
command line tool.
Test case 599 was created and it was used to repeat the bug and then
verify the fix.
Bug: http://curl.haxx.se/bug/view.cgi?id=3510057
Reported by: Michael Wallner
When CURLOPT_REFERER has been used, curl_easy_reset() did not properly
clear it.
Verified with the new test 598
Bug: http://curl.haxx.se/bug/view.cgi?id=3481551
Reported by: Michael Day
591 -> FTP multi PORT and 425 on upload
592 -> FTP multi PORT and 421 on upload
593 -> FTP multi PORT upload, no data conn and no transient neg. reply
594 -> FTP multi PORT upload, no data conn and no positive prelim. reply
1206 -> FTP PORT and 425 on download
1207 -> FTP PORT and 421 on download
1208 -> FTP PORT download, no data conn and no transient negative reply
1209 -> FTP PORT download, no data conn and no positive preliminary reply
This test is created to verify Rene Bernhardt's patch which makes sure
libcurl properly _not_ deals with Negotiate if not asked to even if the
proxy says it can serve it.
When doing a multipart formpost with a read callback, and that callback
returns CURL_READFUNC_ABORT, that return code must be properly
propagated back and handled accordingly. Previously it would be handled
as a zero byte read which would cause a hang!
Added test case 587 to verify. It uses the lib554.c source code with a
small ifdef.
Reported by: Anton Bychkov
Bug: http://curl.haxx.se/mail/lib-2011-10/0097.html
This test case is meant to verify that the logic in commit
60172a0446 actually works. This test failed for me before that
change and it works after it.
Add test 582 for uploading a file using sftp and the multi interface.
(Patch and test slightly tweaked by Daniel Stenberg)
Initially marked as disabled until it is fixed in the source.
Test 580 is removed again for two reasons:
1) Some compilers aren't satisfied by just a data variable called 'test'
when first.o wants a function called 'test'. The Solaris compiler says
"ld: warning: symbol `test' has differing types:" while the AIX compiler
downright rejects it.
2) Test case 1119 that was added after this test is way more complete
and cover everything test 580 does and more without introducing the same
problems.
The new perl script mk580.pl generates a C table in a fresh source file
named lib580.c and if that compiles fine we know that the file
docs/libcurl/symbols-in-versions at least doesn't include any symbols
that are misspelled.
An additional feature would be to somehow scan curl/curl.h and compare
with symbols-in-versions to see if there are symbols missing.
The 66 bytes checked are those 38 bytes with the chunked encoding
headers added: 8+8+10+35+5 = 66
The three-letter words become 8 bytes on the wire because they are sent
like: "3\r\none\r\n"
... and there's the trailing 5 bytes write after the four lines since
the final chunk is sent (which is "0\r\n\r\n").
A shared library tests/libtest/.libs/lihostname.so is preloaded in NTLM
test-cases to override the system implementation of gethostname(). It
makes it possible to test the NTLM authentication for exact match, and
this way test the implementation of MD4 and DES.
If LD_PRELOAD doesn't work, a debug build willl also workk as debug
builds are now made to prefer a specific environment variable and will
then return that content as host name instead of the actual one.
Kamil wrote the bulk of this, Daniel Stenberg polished it.
Dirk Manske reported a regression. When connecting with the multi
interface, there were situations where libcurl wouldn't store
connect time correctly as it used to (and is documented to) do.
Using his fine sample program we could repeat it, and I wrote up
test case 573 using that code. The problem does not easily show
itself using the local test suite though.
The fix, also as suggested by Dirk, is a bit on the ugly side as
it adds yet another call to Curl_verboseconnect() and setting the
TIMER_CONNECT time. That situation is subject for some closer
inspection in the future.
POST using a read callback, with Digest authentication and
"Transfer-Encoding: chunked" enforced. I would then cause the first request
to be wrongly sent and then basically hang until the server closed the
connection. I fixed the problem and added test case 565 to verify it.
If the CURLOPT_PORT option is used on an FTP URL like
"ftp://example.com/file;type=A" the ";type=A" is stripped off.
I added test case 562 to verify, only to find out that I couldn't repeat
this bug so I hereby consider it not a bug anymore!
