similar to: [PATCH] lib: Don't abort if a signal handler calls exit(2) during a guestfs_* function.

Since each ACQUIRE_LOCK/RELEASE_LOCK call must balance, this code is difficult to debug. Enable DEBUG_LOCK to add some prints which can help. The only definitive list of public APIs is found indirectly in the generator (in generator/c.ml : globals). --- generator/c.ml | 18 ++++++++++++++ src/errors.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++---- src/events.c

v2 -> v3: - Lots of code cleanups. - Documentation. However this is still spooling the file into a temporary before the upload. It turns out that fixing this is going to require a small change to qemu. Rich.

Hi, On Saturday 06 June 2015 14:20:38 Richard W.M. Jones wrote: > Since each ACQUIRE_LOCK/RELEASE_LOCK call must balance, this code is > difficult to debug. Enable DEBUG_LOCK to add some prints which can > help. There's some way this could be simplified: > const char * > guestfs_last_error (guestfs_h *g) > { > - return g->last_error; > + const char *r; >

The included patch for openssh 3.0.2p1 implements overlapping read requests for the sftp client. It should be able to handle weird cases such as shriking files and reordered responses. This is only the first shot, and I'd be happy for any comments. I plan to implement something similar for the write path if this works out well. The maximum number of outstanding requests is quite high at

We permit the following constructs in libguestfs code: if (guestfs_some_call (g) == -1) { fprintf (stderr, "failed: error is %s\n", guestfs_last_error (g)); } and: guestfs_push_error_handler (g, NULL, NULL); guestfs_some_call (g); guestfs_pop_error_handler (g); Neither of these would be safe if we allowed the handle to be used from threads concurrently, since the error

We permit the following constructs in libguestfs code: if (guestfs_some_call (g) == -1) { fprintf (stderr, "failed: error is %s\n", guestfs_last_error (g)); } and: guestfs_push_error_handler (g, NULL, NULL); guestfs_some_call (g); guestfs_pop_error_handler (g); Neither of these would be safe if we allowed the handle to be used from threads concurrently, since the error

- fix the syntax of hyperlinks - replace the deprecation text with a simplier named list item, so it's more visible - use a named list item for the pointer to the C documentation of each API - add a named list item for the version of each API --- generator/ruby.ml | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-) diff --git a/generator/ruby.ml

Introduces private pool of muteces (per_handle_locks) guarded by read-write mutex (locks_lock) using guestfs__per_handle_lock_lock(g) and guestfs__per_handle_lock_unlock(g), where g is current valid guestfs_h handle. Above two are used in generator/c.ml at begin and end of each non-deamon call. New lock can be created using guestfs__per_handle_lock_add(g) and destroyed with

From: Sebastian Andrzej Siewior <sebastian at breakpoint.cc> The "zstd at breakpoint.cc" compression algorithm enables ZSTD based compression as defined in RFC8478. The compression is delayed until the server sends the SSH_MSG_USERAUTH_SUCCESS which is the same time as with the "zlib at openssh.com" method. Signed-off-by: Sebastian Andrzej Siewior <sebastian at

--- ../openssh/sftp-client.c Fri Feb 9 08:44:24 2001 +++ sftp-client.c Fri Feb 9 19:14:01 2001 @@ -331,7 +331,7 @@ error("Couldn't read directory: %s", fx2txt(status)); do_close(fd_in, fd_out, handle, handle_len); - return(NULL); + return(0); } } else if (type != SSH2_FXP_NAME) fatal("Expected SSH2_FXP_NAME(%d) packet, got %d",

https://bugzilla.mindrot.org/show_bug.cgi?id=2171 Bug ID: 2171 Summary: potential fd leak Product: Portable OpenSSH Version: -current Hardware: Other OS: Linux Status: NEW Severity: enhancement Priority: P5 Component: sftp Assignee: unassigned-bugs at mindrot.org

https://bugzilla.redhat.com/show_bug.cgi?id=769680

This setOldClass(c("file", "connection")) .A <- setRefClass("A", fields=list(con="connection"), methods=list( finalize = function() { if (isOpen(con)) close(con) })) f <- tempdir() a <- .A$new(con=file(f, "rb")) close(a$con) a <- .A$new(con=file(f,

https://bugzilla.mindrot.org/show_bug.cgi?id=1375 Summary: sftp-client leaks handles on failure to open local file Classification: Unclassified Product: Portable OpenSSH Version: 4.7p1 Platform: All OS/Version: All Status: NEW Severity: minor Priority: P2 Component: sftp AssignedTo: bitbucket

The API for UNIX sockets limits the path to a static size (usually 104 or 108 characters, NULL included), which is internally represented as UNIX_PATH_MAX. If the temporary directory set is long enough (e.g. when running tools as uninstalled, using "run") then these socket paths get trucated, if not even cause failures when binding the sockets. Introduce a new internal API to create

Again, this has been lightly tested. I think there still are a few glitchs. 1. stole progressmeter() from scp.c - clean up and simplified a little to remove the 'flag' status. It now understands how to initialize itself and how to terminate itself. Along with a malloced status bar instead of the original fix width bar. 2. removed all initialization code from scp.c for progressmeter()

Well, I get these error messages during the execution of guestfs_mount_local_run(). What can I do then ? On Wed, Jul 17, 2013 at 5:08 PM, Richard W.M. Jones <rjones@redhat.com>wrote: > On Wed, Jul 17, 2013 at 04:43:34PM +0300, Or Goshen wrote: > > Hi, > > > > When I register a callback for events with this function call: > > eh = guestfs_set_event_callback(g,

Introduce a new read-only API to get a path where to store temporary sockets: this is different than tmpdir, as we need short paths for sockets (due to sockaddr_un::sun_path), and is either XDG_RUNTIME_DIR if set, or /tmp. Adapt guestfs_int_create_socketname to create sockets in that location, checking whether the resulting path still fits in the limited buffer. Furthermore, print sockdir and

Introduce a new read-only API to get a path where to store temporary sockets: this is different from tmpdir, as we need short paths for sockets (due to sockaddr_un::sun_path), and it is either XDG_RUNTIME_DIR if set, or /tmp; adapt guestfs_int_create_socketname to create sockets in that location. Furthermore, print sockdir and XDG_RUNTIME_DIR in test-tool for debugging. --- fish/guestfish.pod

Previous discussion here: https://www.redhat.com/archives/libguestfs/2015-June/thread.html#00048 v2: - Use a cleanup handler to release the lock. - Rebase to upstream. Note I have not fixed the problem(s) with error handling (patch 3).