362 Commits

Author SHA1 Message Date
Daniel Stenberg
ad13de92d4 Fix _libssh2_channel_write(): fixed the write loop to also read from the
transport layer better, to avoid draining the window size when sending large
packets. I also fixed the return code for it to return the number of bytes
handled in this single invoke (and not the cumulative amount).
2009-05-20 14:34:30 +02:00
Daniel Stenberg
b755f3eb10 First, only try to send max packet size *4 to channel write, as trying to send
much larger packages only cause internal problems and much larger allocations.

Also fix sftp_write() when _libssh2_channel_write() returns that a packet was
only partially sent as that is not an error.

Fixed a few error messages to more accurately point out the problem
2009-05-20 14:31:46 +02:00
Daniel Stenberg
b95fe985af If recv() returns something less than the full buffer we know we're done for
now, so skip looping and doing another read that then simply will cause an
EAGAIN as it may trick the user into believing things that aren't true.
2009-05-20 14:30:05 +02:00
Daniel Stenberg
18e03739be typo 2009-05-12 10:10:55 +00:00
Daniel Stenberg
e367e1ad4e comment typo and some indent fixes 2009-05-12 10:07:30 +00:00
Daniel Stenberg
f7149d1830 Partial fix based on the patch and discussions in #2788319 2009-05-12 09:33:22 +00:00
Daniel Stenberg
042b545247 clear the passed in struct when the entry has been removed to help prevent
mistakes due to re-use
2009-05-12 09:32:51 +00:00
Daniel Stenberg
d3542e6f7d split out the knownhost code from hostkey.c into its own separate source file
now: knownhost.c
2009-05-07 17:21:56 +00:00
Daniel Stenberg
0cee15dc0c this was never supposed to be added for real 2009-05-07 17:21:14 +00:00
Daniel Stenberg
4b991b232d My knownhost work as of right now. It works at least partly. More tests and
tweaks will come.
2009-05-07 13:09:48 +00:00
Daniel Stenberg
9412588373 bug #2785173 pointed out that we really must call _libssh2_transport_read()
in loops until it returns < 0 when we call it, as if we just call it once we
may drain the socket for data and then leave unused in-memory data that we
won't detect because the socket is back to idle...
2009-05-01 19:07:20 +00:00
Daniel Stenberg
9f104cd883 Markus posted a bug report about a bad 0-return from libssh2_channel_read:
http://libssh2.haxx.se/mail/libssh2-devel-archive-2009-04/0076.shtml

  And it was indeed a bad loop that terminated too early due to a receveived
  close packet.
2009-04-30 10:30:26 +00:00
Daniel Stenberg
e63cbd907f Jussi Mononen pointed out we used an unsigned variable to store negative
values in, when a plain int is much better anyway since it matches the return
type.
2009-04-06 14:50:29 +00:00
Daniel Stenberg
74f4b7817b Uh, it is important that we don't set the local state 'local.close' to TRUE
until _after_ we're done as otherwise we might not get called again properly
to clean the entire thing since the close state is checked in
libssh2_channel_free
2009-03-31 08:52:25 +00:00
Daniel Stenberg
9ba11a374c Removed the trace output that said "May block until packet of type..." since
it simply wasn't true for non-blocking operations.
2009-03-30 09:19:49 +00:00
Daniel Stenberg
117b95e6e3 made sure the SFTP code makes more use of internal non-blocking functions
instead of the external API entries
2009-03-29 22:09:29 +00:00
Daniel Stenberg
60ee30b6c9 Fixed the memory leak in in libssh2_userauth_hostbased_fromfile_ex() that
Jean-Louis Charton found. Additional minor indent and comment edits.
2009-03-28 21:31:53 +00:00
Daniel Stenberg
55744f8f34 indent changes and removed libssh2_ prefix from static function 2009-03-28 21:22:16 +00:00
Yang Tse
8f78a935ca include libssh2_priv.h instead of libssh2.h in order to get
LIBSSH2_API properly defined on LIBSSH2_WIN32 build targets
2009-03-28 12:50:34 +00:00
Daniel Stenberg
75d5830490 Additional non-blocking fixes. 2009-03-27 22:24:09 +00:00
Daniel Stenberg
103a6e9e0b Me and my long fingers couldn't resist "merging" these statements in a previous
commit but of course that broke functionality and this is a revert of that
change.
2009-03-27 21:52:49 +00:00
Daniel Stenberg
76dab196b1 Dump to stderr simply because the other tracing functions in misc.c already
use stderr and mixing output "channels" causes too much confusion for me!
2009-03-27 21:51:43 +00:00
Daniel Stenberg
4353711127 fixed sftp_shutdown() for the blockiness and fixed libssh2_channel_close()
to work properly non-blocking
2009-03-27 20:20:48 +00:00
Daniel Stenberg
ad8b49b208 Fixed the order of the check in the BLOCK_ADJUST() macro so that it also works
for libssh2_session_free()
2009-03-27 12:48:03 +00:00
Daniel Stenberg
5936d9ce4e Fix the BLOCK_ADJUST_ERRNO() macro to behave properly in blocking mode. 2009-03-27 09:37:48 +00:00
Daniel Stenberg
f390c8e02d fix comments, moved to htonu64 function to sftp.c and made it static 2009-03-27 07:03:00 +00:00
Daniel Stenberg
9df891e412 More cleaning up converting libssh2_ to _libssh2_ for internal functions and
removing libssh2_ from static functions. Added mac.h.
2009-03-26 22:25:23 +00:00
Daniel Stenberg
fc28f33384 comp.c only as a single _libssh2_ function, no external one 2009-03-26 22:09:35 +00:00
Daniel Stenberg
eabe072496 * Renamed the functions in src/transport.c to be _libssh2_transport_ prefixed
and introduced a transport.h header.

* Fixed the blocking mode to only change behavior not the actual underlying
  socket mode so we now always work with non-blocking sockets. This also
  introduces a new rule of thumb in libssh2 code: we don't call the
  external function calls internally. We use the internal (non-blocking)
  ones!

* libssh2_channel_receive_window_adjust2 was added and
  libssh2_channel_receive_window_adjust is now deprecated

* Introduced "local" header files with prototypes etc for different parts
  instead of cramming everything into libssh2_priv.h. channel.h is the
  first.
2009-03-26 15:41:14 +00:00
Daniel Stenberg
2e548c1ca0 Updated comments and removed "libssh2_" prefixes from static functions 2009-03-23 13:35:20 +00:00
Daniel Stenberg
3d97a87669 ENOTCONN obviously doesn't exist in win32 so don't try to map that 2009-03-23 13:08:17 +00:00
Daniel Stenberg
022cac8510 [no comment] 2009-03-23 12:53:50 +00:00
Daniel Stenberg
d187a0fdb6 EEK revert the part of the previous commit that was just my non-functional
attempts to make things work with blocking sockets.
2009-03-23 12:51:53 +00:00
Daniel Stenberg
1936d167d1 I cut out the poll_socket argument from the protos for_libssh2_packet_askv()
and _libssh2_packet_ask() since they are internal and nothing in the code base
uses that functionality == simplicity with no drawbacks!
2009-03-23 12:43:59 +00:00
Daniel Stenberg
26b65e06b1 pass in unsigned ints for u32 values and not longs 2009-03-19 12:54:58 +00:00
Daniel Stenberg
468272d648 - (Mar 19 2009) Daniel Stenberg: based on a patch by "E L" we now use errno
properly after recv() and send() calls (that internally are now known as
  _libssh2_recv() and _libssh2_send()) so that the API and more works fine on
  windows too!
2009-03-19 12:45:59 +00:00
Daniel Stenberg
53ad07633e re-indented for clarity 2009-03-17 14:58:24 +00:00
Daniel Stenberg
0b7f20566c re-indented _libssh2_channel_packet_data_len() somewhat to make the huge
condition somewhat easier on the eye
2009-03-17 14:44:38 +00:00
Daniel Stenberg
cc5e952fa0 A cleanup effort: libssh2_ prefixes only on external APIs. Use _libssh2_ prefix
for library-wide internal functions. Don't use any of those on static functions.
I also did some comments and whitespace changes.
2009-03-17 13:48:35 +00:00
Simon Josefsson
c7a1d30590 Use libtool -export-symbols-regex to reduce numer of exported symbols. 2009-03-17 10:19:54 +00:00
Daniel Stenberg
94090cf55c Patch from bug report #1960894 by "Heiner" applied, and edited by me. It makes
SCP recv and send deal with file names containing spaces (and other
shell-special letters) by quoting them before they are passed to the remote
server.
2009-03-15 23:28:25 +00:00
Daniel Stenberg
ce9d44a3ff fixed some wrong comments and did a minor code edit and some further
whitespace changes
2009-03-15 23:07:09 +00:00
Daniel Stenberg
0812744424 fix a compiler warning 2009-03-15 22:40:11 +00:00
Daniel Stenberg
f56b8ef605 simplified how the "scp" command line is created in the libssh2_scp_recv()
function
2009-03-15 22:38:47 +00:00
Daniel Stenberg
1a630c5c09 Simplified libssh2_channel_read_ex() and made it send window adjustments less
frequent, use a few less struct fields in the channel struct and improved
reading from the network with libssh2_packet_read(). I also modified the
windowing algorithm and now use a much larger window. This greatly enhances
SSH/SCP performance. I also increased the size of the buffer the transport
layer uses from 4k to 16K.
2009-03-15 22:29:23 +00:00
Daniel Stenberg
3793b6768d oops, use %06d for the microseconds! 2009-03-14 23:04:25 +00:00
Daniel Stenberg
269bd5e9a4 re-indented and other whitespace edits 2009-03-14 22:29:48 +00:00
Daniel Stenberg
d7566e5065 improved trace logs with time stamps 2009-03-14 22:29:28 +00:00
Daniel Stenberg
a52f5d3b7e Cleanups, that do seem to have boosted SFTP download performance up to 300% in
some tests:

  * cut off "_ex" from several internal function names

  * corrected some log outputs

  * simplified libssh2_channel_read_ex() and made it much faster in the process

  * cut out {{{ and }}} comments that were incorrect anyway

  * fixed sftp_packet_ask() to return the correct packet by using memcmp() and
    not strncmp()

  * fixed mkdir()'s wait for packet to use the correct request_id - it
    semi-worked previously because strncmp() in sftp_packet_ask() made it
    match far too easily.

  * took away the polling functionality from sftp_packet_ask() since it wasn't
    used
2009-03-13 22:15:27 +00:00
Daniel Stenberg
5f32304805 - (Mar 7 2009) Olivier Hervieu pointed out a flaw in the
libssh2_channel_x11_req_ex() function that made it produce a crappy random
  chunk of data. Peter Stuge improved the fix to not do out-of-boundary
  writes. I (Daniel Stenberg) replaced the snprintf() with a plain sprintf()
  since the size argument wasn't adding anything anyway.
2009-03-07 22:08:05 +00:00