A recent change in glibc 2.42.9000 [1] changes the return type of
strstr() and other string functions to be 'const char *' when the
input is a 'const char *'.
This breaks the build in various files with errors such as :
error: initialization discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]
208 | char *pidstr = strstr(filename, "%");
| ^~~~~~
Fix this by changing the type of the variables that store the result
of these functions to 'const char *'.
[1] https://sourceware.org/git/?p=glibc.git;a=commit;h=cd748a63ab1a7ae846175c532a3daab341c62690
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20251209174328.698774-1-clg@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
The code checks existance of a command (halt/poweroff/reboot) by using
stat(2) and immediately checking for S_ISLNK() on the returned stat
struct. This check will never be true, because stat(2) always follows
symbolic links and hence will either return ENOENT (in case of dangling
symlink) or the properties for the final target file. It is lstat(2)
which might return information about the symlink itself. However, even
there, we want to check the final file properties, not the first symlink.
This check - S_ISLNK - is harmful but useless in this case. However, it
is confusing and it helps the wrong usage of stat(2) to spread, so it is
better to remove it.
Additionally, the code would better to check for the executable bits
of the final file, not check if it's a regular file - it's sort of
dubious to have anything but regular files in /sbin/.
But a POSIX system provides another command which suits the purpose
perfectly: it is access(2). And it is so simple that it's not
necessary to create a separate function when usin it.
Replace stat(2) with access(X_OK) to check for file existance in
qga/commands-posix.c
Fixes: c5b4afd4d5 "qga: Support guest shutdown of BusyBox-based systems"
Reviewed-by: Rodrigo Dias Correa <r@drigo.nl>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Solaris declares getloadavg() in <sys/loadavg.h>:
getloadavg(3C) Standard C Library Functions getloadavg(3C)
NAME
getloadavg - get system load averages
SYNOPSIS
#include <sys/loadavg.h>
int getloadavg(double loadavg[], int nelem);
[...]
Oracle Solaris 11.4 23 Jul 2020 getloadavg(3C)
Include it in order to avoid:
../qga/commands-posix.c: In function 'qmp_guest_get_load':
../qga/commands-posix.c:1408:9: error: implicit declaration of function 'getloadavg' [-Wimplicit-function-declaration]
1408 | if (getloadavg(loadavg, G_N_ELEMENTS(loadavg)) < 0) {
| ^~~~~~~~~~
../qga/commands-posix.c:1408:9: warning: nested extern declaration of 'getloadavg' [-Wnested-externs]
../configure relevant output:
C compiler for the host machine: gcc (gcc 14.2.0 "gcc (GCC) 14.2.0")
C linker for the host machine: gcc ld.solaris 5.11-1.3315
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Acked-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20251117203834.83713-2-philmd@linaro.org>
Sweep the entire documentation again. Last done in commit
7270819384 (qga/qapi-schema: Refill doc comments to conform to
current conventions).
To check the generated documentation does not change, I compared the
generated HTML before and after this commit with "wdiff -3". Finds no
differences. Comparing with diff is not useful, as the reflown
paragraphs are visible there.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20251103082354.3273027-5-armbru@redhat.com>
On POSIX systems, the QEMU Guest Agent uses /sbin/shutdown to implement
the command guest-shutdown. Systems based on BusyBox, such as Alpine
Linux, don't have /sbin/shutdown. They have instead three separate
commands: poweroff, reboot, and halt.
Change the QEMU Guest Agent to, depending on the mode argument, use
/sbin/{poweroff,halt,reboot} when they exist, falling back to
/sbin/shutdown when they don't.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2589
Signed-off-by: Rodrigo Dias Correa <r@drigo.nl>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250926214015.120338-1-r@drigo.nl
Signed-off-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Instead of open-coded g_unix_set_fd_nonblocking() calls, use
QEMU wrapper qemu_set_blocking().
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
[DB: fix missing closing ) in tap-bsd.c, remove now unused GError var]
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Use common qemu_set_blocking() instead.
Note that pre-patch the behavior of Win32 and Linux realizations
are inconsistent: we ignore failure for Win32, and assert success
for Linux.
How do we convert the callers?
1. Most of callers call qemu_socket_set_nonblock() on a
freshly created socket fd, in conditions when we may simply
report an error. Seems correct switching to error handling
both for Windows (pre-patch error is ignored) and Linux
(pre-patch we assert success). Anyway, we normally don't
expect errors in these cases.
Still in tests let's use &error_abort for simplicity.
What are exclusions?
2. hw/virtio/vhost-user.c - we are inside #ifdef CONFIG_LINUX,
so no damage in switching to error handling from assertion.
3. io/channel-socket.c: here we convert both old calls to
qemu_socket_set_nonblock() and qemu_socket_set_block() to
one new call. Pre-patch we assert success for Linux in
qemu_socket_set_nonblock(), and ignore all other errors here.
So, for Windows switch is a bit dangerous: we may get
new errors or crashes(when error_abort is passed) in
cases where we have silently ignored the error before
(was it correct in all such cases, if they were?) Still,
there is no other way to stricter API than take
this risk.
4. util/vhost-user-server - compiled only for Linux (see
util/meson.build), so we are safe, switching from assertion to
&error_abort.
Note: In qga/channel-posix.c we use g_warning(), where g_printerr()
would actually be a better choice. Still let's for now follow
common style of qga, where g_warning() is commonly used to print
such messages, and no call to g_printerr(). Converting everything
to use g_printerr() should better be another series.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
QGA-VSS writes error using error_setg_win32_internal,
which call g_win32_error_message.
g_win32_error_message - translate a Win32 error code
(as returned by GetLastError()) into the corresponding message.
In the same time, we call error_setg_win32_internal with
error codes from different Windows componets like VSS or
Performance monitor that provides different codes and
can't be converted with g_win32_error_message. In this
case, the empty suffix will be returned so error will be
masked.
This commit directly add hex value of error code.
Reproduce:
- Run QGA command: {"execute": "guest-fsfreeze-freeze-list", "arguments": {"mountpoints": ["D:"]}}
QGA error example:
- before changes:
{"error": {"class": "GenericError", "desc": "failed to add D: to snapshot set: "}}
- after changes:
{"error": {"class": "GenericError", "desc": "failed to add D: to snapshot set: Windows error 0x8004230e: "}}
Reviewed-by: Yan Vugenfirer <yvugenfi@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250825135311.138330-1-kkostiuk@redhat.com
Signed-off-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
In this function we could have this variable not initialized. If this
could be acceptable on error, the variable could be left not initialized
f.e. as follows:
void requester_freeze(int *num_vols, void *mountpoints, ErrorSet *errset)
{
...
if (mountpoints) {
...
if (num_mount_points == 0) {
/* If there is no valid mount points, just exit. */
goto out;
}
}
...
if (!mountpoints) {
...
if (num_fixed_drives == 0) {
goto out; /* If there is no fixed drive, just exit. */
}
}
...
}
Stay on safe side, initialize the variable at the beginning.
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Kostiantyn Kostiuk <kkostiuk@redhat.com>
CC: Michael Roth <michael.roth@amd.com>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250807133221.1135453-1-den@openvz.org
Signed-off-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
When compiling QEMU with --enable-ubsan there is a undefined behavior
warning when running "make check":
.../qga/commands-linux.c:452:15: runtime error: applying non-zero offset 5 to null pointer
#0 0x55ea7b89450c in build_guest_fsinfo_for_pci_dev ..../qga/commands-linux.c:452:15
Fix it by avoiding the additional pointer variable here and use an
"offset" integer variable instead.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250730072709.27077-1-thuth@redhat.com
Signed-off-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Commit v9.0.0-343-g2048129625 introduced usage of
g_file_set_contents() function to write to /sys/power/state.
This function uses G_FILE_SET_CONTENTS_CONSISTENT flag to
g_file_set_contents_full(), which is implemented by creating
a temp file in the same directory and renaming it to the final
destination. Which is not how sysfs works.
Here, there's not a big deal to do open/write/close - it becomes
almost the same as using g_file_set_contents[_full](). But it
does not have surprises like this.
Also, since this is linux code, it should be ok to use %m in
the error reporting function.
Fixes: 2048129625 "qga/commands-posix: don't do fork()/exec() when suspending via sysfs"
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/3057
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250801115316.6845-1-mjt@tls.msk.ru>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
"Returns: <description>" is rendered like
Return: <Type> – <description>
Mentioning the type in the description again is commonly redundant.
There is just one such description. Rephrase it not to mention the
type.
We did this for qapi/ in commit f7296f8de5 (qapi: rephrase return
docs to avoid type name).
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20250717115246.3830007-4-armbru@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
The QAPI doc generator recently started to auto-generate return
documentation when there is no "Returns:" section (commit 636c96cd77
"qapi: Fix undocumented return values by generating something").
Remove "Returns:" sections where the auto-generated text is obviously
no worse. For instance, guest-info's documentation changes from
Return:
GuestAgentInfo -- GuestAgentInfo
to
Return:
GuestAgentInfo
The auto-generated returns all are in the exact same spot.
We did this for qapi/ in commit 0462da9d6b (qapi: remove trivial
"Returns:" sections).
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20250717115246.3830007-3-armbru@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Tagged sections are only recognized at the beginning of a paragraph.
guest-network-get-route's Returns: isn't, and therefore gets rendered
as ordinary text within its paragraph:
Retrieve information about route of network. Returns: List of route
info of guest.
Since there is no (recognized) Returns: section, the doc generator
adds
Return:
[GuestNetworkRoute]
Note: only since recent commit 636c96cd77 (qapi: Fix undocumented
return values by generating something).
Insert the required blank line so that Returns: is recognized.
Result:
Retrieve information about route of network.
Return:
[GuestNetworkRoute] -- List of route info of guest.
Fixes: commit 8e326d36dd (qga/linux: Add new api 'guest-network-get-route')
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20250717115246.3830007-2-armbru@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
This commit improves the QGA VSS provider installation flow by attempting to unregister the VSS provider if it's already
found during installation. This allows for a retry of installation even if a previous unregistration failed or was not performed.
This will prevent inconsistencies between QGA and QGA-VSS versions.
Before this commit, QGA can use QGA-VSS from the previous installation.
Signed-off-by: Elizabeth Ashurov <eashurov@redhat.com>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Message-ID: <20250618091806.170110-1-eashurov@redhat.com>
Signed-off-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
