search for: drive_serv

Note this patch requires a non-upstream qemu patch that I've been experimenting with. See qemu-devel list. Rich.

On Tuesday 11 March 2014 23:13:46 Richard W.M. Jones wrote: > diff --git a/src/drives.c b/src/drives.c > index 2c85b52..68e37f7 100644 > --- a/src/drives.c > +++ b/src/drives.c > @@ -115,7 +115,8 @@ static struct drive * > create_drive_file (guestfs_h *g, const char *path, > bool readonly, const char *format, > const char *iface,

This series improves ceph rbd support in libguestfs. It uses the servers list, adds support for a custom username, and starts to add support for custom secret.

...p;& c != ']') - return 0; - } - return 1; -} +#define VALID_HOSTNAME(str) \ + guestfs_int_string_is_valid ((str), 1, 255, true, true, "-.:[]") /** * Check the port number is reasonable. @@ -700,7 +657,7 @@ parse_one_server (guestfs_h *g, const char *server, struct drive_server *ret) return -1; } free (port_str); - if (!valid_hostname (hostname)) { + if (!VALID_HOSTNAME (hostname)) { error (g, _("invalid hostname '%s'"), hostname); free (hostname); return -1; @@ -711,7 +668,7 @@ parse_one_server (guestfs_h *g...

As discussed in the thread on validating $TERM, it would be good to have a common function to do this. This is such a function. The main advantage is it includes unit tests which the previous functions did not. Rich.

One interesting omission is that we don't allow sockets in the abstract namespace. The API won't let you pass these sockets because they contain a '\0' character in the middle of the string. Therefore this patch doesn't need to deal with those. However we should in future allow that, probably using the '@' character to stand in for the NUL byte, as is used in a few

...aunch-direct.c index e6ed54a..00f3190 100644 --- a/src/launch-direct.c +++ b/src/launch-direct.c @@ -1189,7 +1189,8 @@ qemu_escape_param (guestfs_h *g, const char *param) static char * make_uri (guestfs_h *g, const char *scheme, const char *user, const char *password, - struct drive_server *server, const char *path) + struct drive_server *server, const char *path, + const char *querystring) { xmlURI uri = { .scheme = (char *) scheme, .user = (char *) user }; @@ -1217,6 +1218,7 @@ make_uri (guestfs_h *g, const char *scheme, const char *user,...

QEMU upstream has broken snapshot=on ... again. These two patches stop using it entirely. Instead we run 'qemu-img create' to create overlay disks as required. Note that the libvirt and UML backends were already doing this: The libvirt backend because <transient/> has never worked, and the UML backend was running uml_mkcow because the UML-equivalent syntax of snapshot=on was

.../drives.c @@ -28,6 +28,7 @@ #include <stdbool.h> #include <string.h> #include <unistd.h> +#include <limits.h> #include <netdb.h> #include <arpa/inet.h> #include <assert.h> @@ -645,7 +646,18 @@ parse_one_server (guestfs_h *g, const char *server, struct drive_server *ret) return -1; } ret->transport = drive_transport_unix; - ret->u.socket = safe_strdup (g, server+5); + + /* libvirt requires sockets to be specified as an absolute path + * (see RHBZ#1588451), and it's probably a good idea anyway to + * check the socket...

...h-direct.c index bb06c9e..96ae180 100644 --- a/src/launch-direct.c +++ b/src/launch-direct.c @@ -1207,12 +1207,14 @@ qemu_escape_param (guestfs_h *g, const char *param) static char * make_uri (guestfs_h *g, const char *scheme, const char *user, + const char *password, struct drive_server *server, const char *path) { xmlURI uri = { .scheme = (char *) scheme, .user = (char *) user }; CLEANUP_FREE char *query = NULL; CLEANUP_FREE char *pathslash = NULL; + CLEANUP_FREE char *userauth = NULL; /* Need to add a leading '/' to URI paths since xml...

v1 -> v2: - Rebase on top of Pino's version work. Two patches went upstream, these are the two remaining patches. Note the generation number is still inside the qemu.stat file. We could put it in the filename, I have no particular preference. Rich.

...n 1; } +/** + * Check the block size is reasonable. It can't be other then 512 or 4096. + */ +static int +valid_blocksize (int blocksize) +{ + if (blocksize == 512 || blocksize == 4096) + return 1; + return 0; +} + static int parse_one_server (guestfs_h *g, const char *server, struct drive_server *ret) { @@ -767,6 +788,10 @@ guestfs_impl_add_drive_opts (guestfs_h *g, const char *filename, optargs->bitmask & GUESTFS_ADD_DRIVE_OPTS_COPYONREAD_BITMASK ? optargs->copyonread : false; + data.blocksize = + optargs->bitmask & GUESTFS_ADD_DRIVE_OPTS_BLOCKSIZE_BITM...

Doing qemu feature detection in the direct backend takes ~100ms because we need to run `qemu -help' and `qemu -devices ?', and each of those interacts with glibc's very slow link loader. Fixing the link loader is really hard. Instead memoize the output of those two commands. This patch series first separates all the code dealing with qemu into a separate module (src/qemu.c) and

...eck the block size is reasonable. It can't be other then 0, 512 or 4096. + */ +static int +valid_blocksize (int blocksize) +{ + if (blocksize == 0 || blocksize == 512 || blocksize == 4096) + return 1; + return 0; +} + static int parse_one_server (guestfs_h *g, const char *server, struct drive_server *ret) { @@ -767,6 +788,10 @@ guestfs_impl_add_drive_opts (guestfs_h *g, const char *filename, optargs->bitmask & GUESTFS_ADD_DRIVE_OPTS_COPYONREAD_BITMASK ? optargs->copyonread : false; + data.blocksize = + optargs->bitmask & GUESTFS_ADD_DRIVE_OPTS_BLOCKSIZE_BITM...

...;)) { - error (g, _("for device '/dev/null', format must be 'raw'")); + error (g, _("for device ‘/dev/null’, format must be ‘raw’")); return NULL; } } else { @@ -653,14 +653,14 @@ parse_one_server (guestfs_h *g, const char *server, struct drive_server *ret) if (match2 (g, server, re_hostname_port, &hostname, &port_str)) { if (sscanf (port_str, "%d", &port) != 1 || !valid_port (port)) { - error (g, _("invalid port number '%s'"), port_str); + error (g, _("invalid port number ‘%s’&q...

libguestfs has used double and triple underscores in identifiers. These aren't valid for global names in C++. (http://stackoverflow.com/a/228797) These large but completely mechanical patches change the illegal identifiers to legal ones. Rich.