Re: [PATCH 1/1] RFC: Add Arm CCA support for getting capability
information and running Realm VM
by Akio Kakuno (Fujitsu)
Hi, all!
I'm adding three test for CCA compatibility:
domaincapstest, qemucapabilitiestest, and qemuxmlconftest.
This is because SEV-SNP added these three tests.
I have three questions regarding these tests:
1. How to add tests to qemuxmlconftest
2. How to create launch-security-cca.xml
3. About the file output with VIR_TEST_REGENERATE_OUTPUT=1
1. How to add tests to qemuxmlconftest
Following the example of launch-security-sev-snp tests, I've done the following.
Is this correct?
(1) Placed the following three files in tests/qemuxmlconfdata:
launch-security-cca.xml
launch-security-cca.aarch64-latest.xml
launch-security-cca.aarch64-latest.args
(2) Added the test processing to qemuxmlconftest.c's mymain() function:
DO_TEST_CAPS_ARCH_LATEST_FULL("launch-security-cca",
"aarch64",
ARG_QEMU_CAPS,
QEMU_CAPS_CCA_GUEST,
QEMU_CAPS_LAST);
2. How to create launch-security-cca.xml
Do I need to handmade this from scratch or is there an automated method?
3. About the file output with VIR_TEST_REGENERATE_OUTPUT=1
I created launch-security-cca.aarch64-latest.* using the method described in
doc/advanced-tests.rst.
And, I created the test data for qemucapabilitiestest and domaincapstest using
the method described in tests/qemucapabilitiesdata/README.rst.
VIR_TEST_REGENERATE_OUTPUT=1 ./qemuxmlconftest
VIR_TEST_REGENERATE_OUTPUT=1 ./domaincapstest
VIR_TEST_REGENERATE_OUTPUT=1 ./qemucapabilitiestest
Can I use the generated file for testing as is?
Because doc/advanced-tests.rst says:
"VERY CAREFULLY to ensure they are correct"
I assume that automatically generated expected values are checked for accuracy.
If correct, they are adopted; otherwise, investigation and remediation are undertaken.
However, due to the lack of explicit documentation, we require confirmation.
Also, it appears to be generating expected values based on the execution environment.
Do we need to worry about variations in execution environments?
For example, executing qemuxmlconftest detects the following comparison errors,
such as with aarch64-virt-minimal.aarch64-latest, etc.
Expect [sandbox on,obsolete=deny,elevateprivileges=deny,spawn=deny,resourcecontrol=deny \
-m]
Actual [m]
Best Regards.
3 weeks, 6 days
AppArmor confinement for qemu:///session VMs
by Andrea Bolognani
An issue was recently reported[1] with running unprivileged VMs
configured to use passt on Debian with AppArmor confinement enabled.
After looking into the situation, I am convinced that AppArmor
confinement never really worked for unprivileged VMs. The whole
mechanism is built around the concept of per-VM profiles that are
dynamically generated and registered, but doing so requires write
access to /etc/apparmor.d/ and in general permissions that
unprivileged libvirt will by design not have.
Of course it's unfortunate that unprivileged VMs would be forced to
miss out on the potential benefits of AppArmor isolation, and even
more unfortunate that passt won't work out of the box for
unprivileged VMs, since those are the ones where it makes the most
sense to use passt in the first place.
Stefano suggested introducing a generic "libvirt-user" profile that
would be attached to unprivileged VMs and would be more liberal than
the one used for privileged VMs, since we wouldn't be able to tailor
it to the specifics of the VM, but would at least prevent the worst
of the abuse; specifically, it would only allow R/W access to files
in the current user's home directory.
Does that sound like a reasonable direction? Any other ideas?
In the meantime, Stefano has posted a workaround[2] that, when
applied to passt's AppArmor profile, would allow these VMs to at
least start.
CC'ing people with AppArmor knowledge for awareness.
[1] https://archives.passt.top/passt-dev/20250129104112.0756df5c@elisabeth/T/#u
[2] https://archives.passt.top/passt-dev/20250205163101.3793658-1-sbrivio@red...
--
Andrea Bolognani / Red Hat / Virtualization
3 weeks, 6 days
[PATCH 1/4] virpci: changed the work with PCI via libpciaccess
by Alexander Shursha
Sponsored by: Future Crew, LLC
Signed-off-by: Alexander Shursha <kekek2(a)ya.ru>
---
src/meson.build | 1 +
src/util/virpci.c | 464 ++++++++++------------------------------
tests/qemuhotplugtest.c | 11 +-
tests/qemumemlocktest.c | 9 +
tests/qemuxmlconftest.c | 9 +
tests/virpcimock.c | 21 +-
6 files changed, 160 insertions(+), 355 deletions(-)
diff --git a/src/meson.build b/src/meson.build
index 9413192a55..39788ac4d7 100644
--- a/src/meson.build
+++ b/src/meson.build
@@ -411,6 +411,7 @@ libvirt_lib = shared_library(
dtrace_gen_objects,
dependencies: [
src_dep,
+ pciaccess_dep
],
link_args: libvirt_link_args,
link_whole: [
diff --git a/src/util/virpci.c b/src/util/virpci.c
index 90617e69c6..3f95fa2b3f 100644
--- a/src/util/virpci.c
+++ b/src/util/virpci.c
@@ -29,6 +29,7 @@
#include <sys/types.h>
#include <sys/stat.h>
#include <unistd.h>
+#include <pciaccess.h>
#ifdef __linux__
# include <sys/utsname.h>
@@ -72,7 +73,7 @@ struct _virPCIDevice {
char *name; /* domain:bus:slot.function */
char id[PCI_ID_LEN]; /* product vendor */
- char *path;
+ struct pci_device *device;
/* The driver:domain which uses the device */
char *used_by_drvname;
@@ -359,121 +360,6 @@ virPCIDeviceGetCurrentDriverNameAndType(virPCIDevice *dev,
}
-static int
-virPCIDeviceConfigOpenInternal(virPCIDevice *dev, bool readonly, bool fatal)
-{
- int fd;
-
- fd = open(dev->path, readonly ? O_RDONLY : O_RDWR);
-
- if (fd < 0) {
- if (fatal) {
- virReportSystemError(errno,
- _("Failed to open config space file '%1$s'"),
- dev->path);
- } else {
- VIR_WARN("Failed to open config space file '%s': %s",
- dev->path, g_strerror(errno));
- }
- return -1;
- }
-
- VIR_DEBUG("%s %s: opened %s", dev->id, dev->name, dev->path);
- return fd;
-}
-
-static int
-virPCIDeviceConfigOpen(virPCIDevice *dev)
-{
- return virPCIDeviceConfigOpenInternal(dev, true, true);
-}
-
-static int
-virPCIDeviceConfigOpenTry(virPCIDevice *dev)
-{
- return virPCIDeviceConfigOpenInternal(dev, true, false);
-}
-
-static int
-virPCIDeviceConfigOpenWrite(virPCIDevice *dev)
-{
- return virPCIDeviceConfigOpenInternal(dev, false, true);
-}
-
-static void
-virPCIDeviceConfigClose(virPCIDevice *dev, int cfgfd)
-{
- if (VIR_CLOSE(cfgfd) < 0) {
- VIR_WARN("Failed to close config space file '%s': %s",
- dev->path, g_strerror(errno));
- }
-}
-
-
-static int
-virPCIDeviceRead(virPCIDevice *dev,
- int cfgfd,
- unsigned int pos,
- uint8_t *buf,
- unsigned int buflen)
-{
- memset(buf, 0, buflen);
- errno = 0;
-
- if (lseek(cfgfd, pos, SEEK_SET) != pos ||
- saferead(cfgfd, buf, buflen) != buflen) {
- VIR_DEBUG("Failed to read %u bytes at %u from '%s' : %s",
- buflen, pos, dev->path, g_strerror(errno));
- return -1;
- }
- return 0;
-}
-
-
-/**
- * virPCIDeviceReadN:
- * @dev: virPCIDevice object (used only to log name of config file)
- * @cfgfd: open file descriptor for device config file in sysfs
- * @pos: byte offset in the file to read from
- *
- * read "N" (where "N" is "8", "16", or "32", and appears at the end
- * of the function name) bytes from a PCI device's already-opened
- * sysfs config file and return them as the return value from the
- * function.
- *
- * Returns the value at @pos in the file, or 0 if there was an
- * error. NB: since 0 could be a valid value, occurrence of an error
- * must be determined by examining errno. errno is always reset to 0
- * before the seek/read is attempted (see virPCIDeviceRead()), so if
- * errno != 0 on return from one of these functions, then either the
- * seek or the read operation failed for some reason. If errno == 0
- * and the return value is 0, then the config file really does contain
- * the value 0 at @pos.
- */
-static uint8_t
-virPCIDeviceRead8(virPCIDevice *dev, int cfgfd, unsigned int pos)
-{
- uint8_t buf;
- virPCIDeviceRead(dev, cfgfd, pos, &buf, sizeof(buf));
- return buf;
-}
-
-static uint16_t
-virPCIDeviceRead16(virPCIDevice *dev, int cfgfd, unsigned int pos)
-{
- uint8_t buf[2];
- virPCIDeviceRead(dev, cfgfd, pos, &buf[0], sizeof(buf));
- return (buf[0] << 0) | (buf[1] << 8);
-}
-
-static uint32_t
-virPCIDeviceRead32(virPCIDevice *dev, int cfgfd, unsigned int pos)
-{
- uint8_t buf[4];
- virPCIDeviceRead(dev, cfgfd, pos, &buf[0], sizeof(buf));
- return (buf[0] << 0) | (buf[1] << 8) | (buf[2] << 16) | (buf[3] << 24);
-}
-
static int
virPCIDeviceReadClass(virPCIDevice *dev, uint16_t *device_class)
{
@@ -499,36 +385,6 @@ virPCIDeviceReadClass(virPCIDevice *dev, uint16_t *device_class)
return 0;
}
-static int
-virPCIDeviceWrite(virPCIDevice *dev,
- int cfgfd,
- unsigned int pos,
- uint8_t *buf,
- unsigned int buflen)
-{
- if (lseek(cfgfd, pos, SEEK_SET) != pos ||
- safewrite(cfgfd, buf, buflen) != buflen) {
- VIR_WARN("Failed to write to '%s' : %s", dev->path,
- g_strerror(errno));
- return -1;
- }
- return 0;
-}
-
-static void
-virPCIDeviceWrite16(virPCIDevice *dev, int cfgfd, unsigned int pos, uint16_t val)
-{
- uint8_t buf[2] = { (val >> 0), (val >> 8) };
- virPCIDeviceWrite(dev, cfgfd, pos, &buf[0], sizeof(buf));
-}
-
-static void
-virPCIDeviceWrite32(virPCIDevice *dev, int cfgfd, unsigned int pos, uint32_t val)
-{
- uint8_t buf[4] = { (val >> 0), (val >> 8), (val >> 16), (val >> 24) };
- virPCIDeviceWrite(dev, cfgfd, pos, &buf[0], sizeof(buf));
-}
-
typedef int (*virPCIDeviceIterPredicate)(virPCIDevice *, virPCIDevice *,
void *);
@@ -610,7 +466,6 @@ virPCIDeviceIterDevices(virPCIDeviceIterPredicate predicate,
*/
static int
virPCIDeviceFindCapabilityOffset(virPCIDevice *dev,
- int cfgfd,
unsigned int capability,
unsigned int *offset)
{
@@ -619,11 +474,13 @@ virPCIDeviceFindCapabilityOffset(virPCIDevice *dev,
*offset = 0; /* assume failure (*nothing* can be at offset 0) */
- status = virPCIDeviceRead16(dev, cfgfd, PCI_STATUS);
+ pci_device_cfg_read_u16(dev->device, &status, PCI_STATUS);
+
if (errno != 0 || !(status & PCI_STATUS_CAP_LIST))
goto error;
- pos = virPCIDeviceRead8(dev, cfgfd, PCI_CAPABILITY_LIST);
+ pci_device_cfg_read_u8(dev->device, &pos, PCI_CAPABILITY_LIST);
+
if (errno != 0)
goto error;
@@ -635,7 +492,9 @@ virPCIDeviceFindCapabilityOffset(virPCIDevice *dev,
* capabilities here.
*/
while (pos >= PCI_CONF_HEADER_LEN && pos != 0xff) {
- uint8_t capid = virPCIDeviceRead8(dev, cfgfd, pos);
+ uint8_t capid;
+ pci_device_cfg_read_u8(dev->device, &capid, pos);
+
if (errno != 0)
goto error;
@@ -646,7 +505,8 @@ virPCIDeviceFindCapabilityOffset(virPCIDevice *dev,
return 0;
}
- pos = virPCIDeviceRead8(dev, cfgfd, pos + 1);
+ pci_device_cfg_read_u8(dev->device, &pos, pos + 1);
+
if (errno != 0)
goto error;
}
@@ -665,7 +525,6 @@ virPCIDeviceFindCapabilityOffset(virPCIDevice *dev,
static unsigned int
virPCIDeviceFindExtendedCapabilityOffset(virPCIDevice *dev,
- int cfgfd,
unsigned int capability)
{
int ttl;
@@ -677,7 +536,7 @@ virPCIDeviceFindExtendedCapabilityOffset(virPCIDevice *dev,
pos = PCI_EXT_CAP_BASE;
while (ttl > 0 && pos >= PCI_EXT_CAP_BASE) {
- header = virPCIDeviceRead32(dev, cfgfd, pos);
+ header = pci_device_cfg_read_u32(dev->device, &header, pos);
if ((header & PCI_EXT_CAP_ID_MASK) == capability)
return pos;
@@ -693,7 +552,7 @@ virPCIDeviceFindExtendedCapabilityOffset(virPCIDevice *dev,
* not have FLR, 1 if it does, and -1 on error
*/
static bool
-virPCIDeviceDetectFunctionLevelReset(virPCIDevice *dev, int cfgfd)
+virPCIDeviceDetectFunctionLevelReset(virPCIDevice *dev)
{
uint32_t caps;
unsigned int pos;
@@ -707,7 +566,7 @@ virPCIDeviceDetectFunctionLevelReset(virPCIDevice *dev, int cfgfd)
* on SR-IOV NICs at the moment.
*/
if (dev->pcie_cap_pos) {
- caps = virPCIDeviceRead32(dev, cfgfd, dev->pcie_cap_pos + PCI_EXP_DEVCAP);
+ pci_device_cfg_read_u32(dev->device, &caps, dev->pcie_cap_pos + PCI_EXP_DEVCAP);
if (caps & PCI_EXP_DEVCAP_FLR) {
VIR_DEBUG("%s %s: detected PCIe FLR capability", dev->id, dev->name);
return true;
@@ -718,11 +577,13 @@ virPCIDeviceDetectFunctionLevelReset(virPCIDevice *dev, int cfgfd)
* the same thing, except for conventional PCI
* devices. This is not common yet.
*/
- if (virPCIDeviceFindCapabilityOffset(dev, cfgfd, PCI_CAP_ID_AF, &pos) < 0)
+ if (virPCIDeviceFindCapabilityOffset(dev, PCI_CAP_ID_AF, &pos) < 0)
goto error;
if (pos) {
- caps = virPCIDeviceRead16(dev, cfgfd, pos + PCI_AF_CAP);
+ uint16_t caps16;
+ pci_device_cfg_read_u16(dev->device, &caps16, pos + PCI_AF_CAP);
+ caps = caps16;
if (caps & PCI_AF_CAP_FLR) {
VIR_DEBUG("%s %s: detected PCI FLR capability", dev->id, dev->name);
return true;
@@ -754,13 +615,13 @@ virPCIDeviceDetectFunctionLevelReset(virPCIDevice *dev, int cfgfd)
* internal reset, not just a soft reset.
*/
static bool
-virPCIDeviceDetectPowerManagementReset(virPCIDevice *dev, int cfgfd)
+virPCIDeviceDetectPowerManagementReset(virPCIDevice *dev)
{
if (dev->pci_pm_cap_pos) {
uint32_t ctl;
/* require the NO_SOFT_RESET bit is clear */
- ctl = virPCIDeviceRead32(dev, cfgfd, dev->pci_pm_cap_pos + PCI_PM_CTRL);
+ pci_device_cfg_read_u32(dev->device, &ctl, dev->pci_pm_cap_pos + PCI_PM_CTRL);
if (!(ctl & PCI_PM_CTRL_NO_SOFT_RESET)) {
VIR_DEBUG("%s %s: detected PM reset capability", dev->id, dev->name);
return true;
@@ -811,26 +672,22 @@ virPCIDeviceIsParent(virPCIDevice *dev, virPCIDevice *check, void *data)
uint8_t header_type, secondary, subordinate;
virPCIDevice **best = data;
int ret = 0;
- int fd;
if (dev->address.domain != check->address.domain)
return 0;
- if ((fd = virPCIDeviceConfigOpenTry(check)) < 0)
- return 0;
-
/* Is it a bridge? */
ret = virPCIDeviceReadClass(check, &device_class);
if (ret < 0 || device_class != PCI_CLASS_BRIDGE_PCI)
- goto cleanup;
+ return ret;
/* Is it a plane? */
- header_type = virPCIDeviceRead8(check, fd, PCI_HEADER_TYPE);
+ pci_device_cfg_read_u8(check->device, &header_type, PCI_HEADER_TYPE);
if ((header_type & PCI_HEADER_TYPE_MASK) != PCI_HEADER_TYPE_BRIDGE)
- goto cleanup;
+ return ret;
- secondary = virPCIDeviceRead8(check, fd, PCI_SECONDARY_BUS);
- subordinate = virPCIDeviceRead8(check, fd, PCI_SUBORDINATE_BUS);
+ pci_device_cfg_read_u8(check->device, &secondary, PCI_SECONDARY_BUS);
+ pci_device_cfg_read_u8(check->device, &subordinate, PCI_SUBORDINATE_BUS);
VIR_DEBUG("%s %s: found parent device %s", dev->id, dev->name, check->name);
@@ -838,8 +695,7 @@ virPCIDeviceIsParent(virPCIDevice *dev, virPCIDevice *check, void *data)
* the direct parent. No further work is necessary
*/
if (dev->address.bus == secondary) {
- ret = 1;
- goto cleanup;
+ return 1;
}
/* otherwise, SRIOV allows VFs to be on different buses than their PFs.
@@ -850,35 +706,26 @@ virPCIDeviceIsParent(virPCIDevice *dev, virPCIDevice *check, void *data)
if (*best == NULL) {
*best = virPCIDeviceNew(&check->address);
if (*best == NULL) {
- ret = -1;
- goto cleanup;
+ return -1;
}
} else {
/* OK, we had already recorded a previous "best" match for the
* parent. See if the current device is more restrictive than the
* best, and if so, make it the new best
*/
- int bestfd;
uint8_t best_secondary;
- if ((bestfd = virPCIDeviceConfigOpenTry(*best)) < 0)
- goto cleanup;
- best_secondary = virPCIDeviceRead8(*best, bestfd, PCI_SECONDARY_BUS);
- virPCIDeviceConfigClose(*best, bestfd);
+ pci_device_cfg_read_u8((*best)->device, &best_secondary, PCI_SECONDARY_BUS);
if (secondary > best_secondary) {
virPCIDeviceFree(*best);
*best = virPCIDeviceNew(&check->address);
if (*best == NULL) {
- ret = -1;
- goto cleanup;
+ return -1;
}
}
}
}
-
- cleanup:
- virPCIDeviceConfigClose(check, fd);
return ret;
}
@@ -902,15 +749,14 @@ virPCIDeviceGetParent(virPCIDevice *dev, virPCIDevice **parent)
*/
static int
virPCIDeviceTrySecondaryBusReset(virPCIDevice *dev,
- int cfgfd,
virPCIDeviceList *inactiveDevs)
{
g_autoptr(virPCIDevice) parent = NULL;
g_autoptr(virPCIDevice) conflict = NULL;
uint8_t config_space[PCI_CONF_LEN];
uint16_t ctl;
- int ret = -1;
- int parentfd;
+ pciaddr_t bytes_read;
+ pciaddr_t bytes_written;
/* Refuse to do a secondary bus reset if there are other
* devices/functions behind the bus are used by the host
@@ -932,8 +778,6 @@ virPCIDeviceTrySecondaryBusReset(virPCIDevice *dev,
dev->name);
return -1;
}
- if ((parentfd = virPCIDeviceConfigOpenWrite(parent)) < 0)
- goto out;
VIR_DEBUG("%s %s: doing a secondary bus reset", dev->id, dev->name);
@@ -941,38 +785,37 @@ virPCIDeviceTrySecondaryBusReset(virPCIDevice *dev,
* for the supplied device since we refuse to do a reset if there
* are multiple devices/functions
*/
- if (virPCIDeviceRead(dev, cfgfd, 0, config_space, PCI_CONF_LEN) < 0) {
+ pci_device_cfg_read(dev->device, config_space, 0, PCI_CONF_LEN, &bytes_read);
+ if (bytes_read < PCI_CONF_LEN) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Failed to read PCI config space for %1$s"),
dev->name);
- goto out;
+ return -1;
}
/* Read the control register, set the reset flag, wait 200ms,
* unset the reset flag and wait 200ms.
*/
- ctl = virPCIDeviceRead16(dev, parentfd, PCI_BRIDGE_CONTROL);
- virPCIDeviceWrite16(parent, parentfd, PCI_BRIDGE_CONTROL,
- ctl | PCI_BRIDGE_CTL_RESET);
+ pci_device_cfg_read_u16(parent->device, &ctl, PCI_BRIDGE_CONTROL);
+
+ pci_device_cfg_write_u16(parent->device, ctl | PCI_BRIDGE_CTL_RESET, PCI_BRIDGE_CONTROL);
g_usleep(200 * 1000); /* sleep 200ms */
- virPCIDeviceWrite16(parent, parentfd, PCI_BRIDGE_CONTROL, ctl);
+ pci_device_cfg_write_u16(parent->device, ctl, PCI_BRIDGE_CONTROL);
g_usleep(200 * 1000); /* sleep 200ms */
- if (virPCIDeviceWrite(dev, cfgfd, 0, config_space, PCI_CONF_LEN) < 0) {
+ pci_device_cfg_write(dev->device, config_space, 0, PCI_CONF_LEN, &bytes_written);
+ if (bytes_written < PCI_CONF_LEN) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Failed to restore PCI config space for %1$s"),
dev->name);
- goto out;
+ return -1;
}
- ret = 0;
- out:
- virPCIDeviceConfigClose(parent, parentfd);
- return ret;
+ return 0;
}
/* Power management reset attempts to reset a device using a
@@ -980,16 +823,19 @@ virPCIDeviceTrySecondaryBusReset(virPCIDevice *dev,
* above we require the device supports a full internal reset.
*/
static int
-virPCIDeviceTryPowerManagementReset(virPCIDevice *dev, int cfgfd)
+virPCIDeviceTryPowerManagementReset(virPCIDevice *dev)
{
uint8_t config_space[PCI_CONF_LEN];
uint32_t ctl;
+ pciaddr_t bytes_read;
+ pciaddr_t bytes_written;
if (!dev->pci_pm_cap_pos)
return -1;
/* Save and restore the device's config space. */
- if (virPCIDeviceRead(dev, cfgfd, 0, &config_space[0], PCI_CONF_LEN) < 0) {
+ pci_device_cfg_read(dev->device, config_space, 0, PCI_CONF_LEN, &bytes_read);
+ if (bytes_read < PCI_CONF_LEN) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Failed to read PCI config space for %1$s"),
dev->name);
@@ -998,20 +844,19 @@ virPCIDeviceTryPowerManagementReset(virPCIDevice *dev, int cfgfd)
VIR_DEBUG("%s %s: doing a power management reset", dev->id, dev->name);
- ctl = virPCIDeviceRead32(dev, cfgfd, dev->pci_pm_cap_pos + PCI_PM_CTRL);
+ pci_device_cfg_read_u32(dev->device, &ctl, dev->pci_pm_cap_pos + PCI_PM_CTRL);
ctl &= ~PCI_PM_CTRL_STATE_MASK;
- virPCIDeviceWrite32(dev, cfgfd, dev->pci_pm_cap_pos + PCI_PM_CTRL,
- ctl | PCI_PM_CTRL_STATE_D3hot);
+ pci_device_cfg_write_u32(dev->device, ctl | PCI_PM_CTRL_STATE_D3hot, dev->pci_pm_cap_pos + PCI_PM_CTRL);
g_usleep(10 * 1000); /* sleep 10ms */
- virPCIDeviceWrite32(dev, cfgfd, dev->pci_pm_cap_pos + PCI_PM_CTRL,
- ctl | PCI_PM_CTRL_STATE_D0);
+ pci_device_cfg_write_u32(dev->device, ctl | PCI_PM_CTRL_STATE_D0, dev->pci_pm_cap_pos + PCI_PM_CTRL);
g_usleep(10 * 1000); /* sleep 10ms */
- if (virPCIDeviceWrite(dev, cfgfd, 0, &config_space[0], PCI_CONF_LEN) < 0) {
+ pci_device_cfg_write(dev->device, config_space, 0, PCI_CONF_LEN, &bytes_written);
+ if (bytes_written < PCI_CONF_LEN) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Failed to restore PCI config space for %1$s"),
dev->name);
@@ -1046,10 +891,10 @@ virPCIDeviceTryPowerManagementReset(virPCIDevice *dev, int cfgfd)
* Always returns success (0) (for now)
*/
static int
-virPCIDeviceInit(virPCIDevice *dev, int cfgfd)
+virPCIDeviceInit(virPCIDevice *dev)
{
dev->is_pcie = false;
- if (virPCIDeviceFindCapabilityOffset(dev, cfgfd, PCI_CAP_ID_EXP, &dev->pcie_cap_pos) < 0) {
+ if (virPCIDeviceFindCapabilityOffset(dev, PCI_CAP_ID_EXP, &dev->pcie_cap_pos) < 0) {
/* an unprivileged process is unable to read *all* of a
* device's PCI config (it can only read the first 64
* bytes, which isn't enough for see the Express
@@ -1065,18 +910,13 @@ virPCIDeviceInit(virPCIDevice *dev, int cfgfd)
* -1), then we blindly assume the most likely outcome -
* PCIe.
*/
- off_t configLen = virFileLength(virPCIDeviceGetConfigPath(dev), -1);
-
- if (configLen != 256)
- dev->is_pcie = true;
-
} else {
dev->is_pcie = (dev->pcie_cap_pos != 0);
}
- virPCIDeviceFindCapabilityOffset(dev, cfgfd, PCI_CAP_ID_PM, &dev->pci_pm_cap_pos);
- dev->has_flr = virPCIDeviceDetectFunctionLevelReset(dev, cfgfd);
- dev->has_pm_reset = virPCIDeviceDetectPowerManagementReset(dev, cfgfd);
+ virPCIDeviceFindCapabilityOffset(dev, PCI_CAP_ID_PM, &dev->pci_pm_cap_pos);
+ dev->has_flr = virPCIDeviceDetectFunctionLevelReset(dev);
+ dev->has_pm_reset = virPCIDeviceDetectPowerManagementReset(dev);
return 0;
}
@@ -1089,7 +929,6 @@ virPCIDeviceReset(virPCIDevice *dev,
g_autofree char *drvName = NULL;
virPCIStubDriver drvType;
int ret = -1;
- int fd = -1;
int hdrType = -1;
if (virPCIGetHeaderType(dev, &hdrType) < 0)
@@ -1114,29 +953,26 @@ virPCIDeviceReset(virPCIDevice *dev,
* be redundant.
*/
if (virPCIDeviceGetCurrentDriverNameAndType(dev, &drvName, &drvType) < 0)
- goto cleanup;
+ return -1;
if (drvType == VIR_PCI_STUB_DRIVER_VFIO) {
VIR_DEBUG("Device %s is bound to %s - skip reset", dev->name, drvName);
ret = 0;
- goto cleanup;
+ return 0;
}
VIR_DEBUG("Resetting device %s", dev->name);
- if ((fd = virPCIDeviceConfigOpenWrite(dev)) < 0)
- goto cleanup;
-
- if (virPCIDeviceInit(dev, fd) < 0)
- goto cleanup;
+ if (virPCIDeviceInit(dev) < 0)
+ return -1;
/* KVM will perform FLR when starting and stopping
* a guest, so there is no need for us to do it here.
*/
if (dev->has_flr) {
ret = 0;
- goto cleanup;
+ return 0;
}
/* If the device supports PCI power management reset,
@@ -1144,11 +980,11 @@ virPCIDeviceReset(virPCIDevice *dev,
* the function, not the whole device.
*/
if (dev->has_pm_reset)
- ret = virPCIDeviceTryPowerManagementReset(dev, fd);
+ ret = virPCIDeviceTryPowerManagementReset(dev);
/* Bus reset is not an option with the root bus */
if (ret < 0 && dev->address.bus != 0)
- ret = virPCIDeviceTrySecondaryBusReset(dev, fd, inactiveDevs);
+ ret = virPCIDeviceTrySecondaryBusReset(dev, inactiveDevs);
if (ret < 0) {
virErrorPtr err = virGetLastError();
@@ -1159,8 +995,6 @@ virPCIDeviceReset(virPCIDevice *dev,
_("no FLR, PM reset or bus reset available"));
}
- cleanup:
- virPCIDeviceConfigClose(dev, fd);
return ret;
}
@@ -1756,28 +1590,6 @@ virPCIDeviceReattach(virPCIDevice *dev,
return 0;
}
-static char *
-virPCIDeviceReadID(virPCIDevice *dev, const char *id_name)
-{
- g_autofree char *path = NULL;
- g_autofree char *id_str = NULL;
-
- path = virPCIFile(dev->name, id_name);
-
- /* ID string is '0xNNNN\n' ... i.e. 7 bytes */
- if (virFileReadAll(path, 7, &id_str) < 0)
- return NULL;
-
- /* Check for 0x suffix */
- if (id_str[0] != '0' || id_str[1] != 'x')
- return NULL;
-
- /* Chop off the newline; we know the string is 7 bytes */
- id_str[6] = '\0';
-
- return g_steal_pointer(&id_str);
-}
-
bool
virPCIDeviceAddressIsValid(virPCIDeviceAddress *addr,
bool report)
@@ -1865,9 +1677,9 @@ virPCIDeviceExists(const virPCIDeviceAddress *addr)
virPCIDevice *
virPCIDeviceNew(const virPCIDeviceAddress *address)
{
+ struct pci_device * device;
+
g_autoptr(virPCIDevice) dev = NULL;
- g_autofree char *vendor = NULL;
- g_autofree char *product = NULL;
dev = g_new0(virPCIDevice, 1);
@@ -1875,31 +1687,21 @@ virPCIDeviceNew(const virPCIDeviceAddress *address)
dev->name = virPCIDeviceAddressAsString(&dev->address);
- dev->path = g_strdup_printf(PCI_SYSFS "devices/%s/config", dev->name);
-
- if (!virFileExists(dev->path)) {
- virReportSystemError(errno,
- _("Device %1$s not found: could not access %2$s"),
- dev->name, dev->path);
+ pci_system_init();
+ device = pci_device_find_by_slot(address->domain, address->bus, address->slot, address->function);
+ if (!device)
return NULL;
- }
-
- vendor = virPCIDeviceReadID(dev, "vendor");
- product = virPCIDeviceReadID(dev, "device");
-
- if (!vendor || !product) {
- virReportError(VIR_ERR_INTERNAL_ERROR,
- _("Failed to read product/vendor ID for %1$s"),
- dev->name);
+ dev->device = g_memdup(device, sizeof(*dev->device));
+ if (!dev->device) {
+ virReportSystemError(errno,
+ _("Not found device domain: %1$d, bus: %2$d, slot: %3$d, function: %4$d"),
+ address->domain, address->bus, address->slot, address->function);
return NULL;
}
-
- /* strings contain '0x' prefix */
- if (g_snprintf(dev->id, sizeof(dev->id), "%s %s", &vendor[2],
- &product[2]) >= sizeof(dev->id)) {
+ if (g_snprintf(dev->id, sizeof(dev->id), "%x %x", dev->device->vendor_id, dev->device->device_id) >= sizeof(dev->id)) {
virReportError(VIR_ERR_INTERNAL_ERROR,
- _("dev->id buffer overflow: %1$s %2$s"),
- &vendor[2], &product[2]);
+ _("dev->id buffer overflow: %1$d %2$d"),
+ dev->device->vendor_id, dev->device->device_id);
return NULL;
}
@@ -1918,10 +1720,10 @@ virPCIDeviceCopy(virPCIDevice *dev)
/* shallow copy to take care of most attributes */
*copy = *dev;
- copy->path = NULL;
- copy->used_by_drvname = copy->used_by_domname = NULL;
+ copy->device = NULL;
+ copy->device = g_memdup(dev->device, sizeof(*dev->device));
+ copy->name = copy->used_by_drvname = copy->used_by_domname = copy->stubDriverName = NULL;
copy->name = g_strdup(dev->name);
- copy->path = g_strdup(dev->path);
copy->used_by_drvname = g_strdup(dev->used_by_drvname);
copy->used_by_domname = g_strdup(dev->used_by_domname);
copy->stubDriverName = g_strdup(dev->stubDriverName);
@@ -1936,10 +1738,11 @@ virPCIDeviceFree(virPCIDevice *dev)
return;
VIR_DEBUG("%s %s: freeing", dev->id, dev->name);
g_free(dev->name);
- g_free(dev->path);
g_free(dev->used_by_drvname);
g_free(dev->used_by_domname);
g_free(dev->stubDriverName);
+ if (dev->device)
+ g_free(dev->device);
g_free(dev);
}
@@ -1971,9 +1774,9 @@ virPCIDeviceGetName(virPCIDevice *dev)
* config file.
*/
const char *
-virPCIDeviceGetConfigPath(virPCIDevice *dev)
+virPCIDeviceGetConfigPath(virPCIDevice *dev G_GNUC_UNUSED)
{
- return dev->path;
+ return NULL;
}
void virPCIDeviceSetManaged(virPCIDevice *dev, bool managed)
@@ -2484,47 +2287,37 @@ virPCIDeviceDownstreamLacksACS(virPCIDevice *dev)
uint16_t flags;
uint16_t ctrl;
unsigned int pos;
- int fd;
- int ret = 0;
uint16_t device_class;
- if ((fd = virPCIDeviceConfigOpen(dev)) < 0)
+ if (virPCIDeviceInit(dev) < 0) {
return -1;
-
- if (virPCIDeviceInit(dev, fd) < 0) {
- ret = -1;
- goto cleanup;
}
if (virPCIDeviceReadClass(dev, &device_class) < 0)
- goto cleanup;
+ return 0;
pos = dev->pcie_cap_pos;
if (!pos || device_class != PCI_CLASS_BRIDGE_PCI)
- goto cleanup;
+ return 0;
- flags = virPCIDeviceRead16(dev, fd, pos + PCI_EXP_FLAGS);
+ pci_device_cfg_read_u16(dev->device, &flags, pos + PCI_EXP_FLAGS);
if (((flags & PCI_EXP_FLAGS_TYPE) >> 4) != PCI_EXP_TYPE_DOWNSTREAM)
- goto cleanup;
+ return 0;
- pos = virPCIDeviceFindExtendedCapabilityOffset(dev, fd, PCI_EXT_CAP_ID_ACS);
+ pos = virPCIDeviceFindExtendedCapabilityOffset(dev, PCI_EXT_CAP_ID_ACS);
if (!pos) {
VIR_DEBUG("%s %s: downstream port lacks ACS", dev->id, dev->name);
- ret = 1;
- goto cleanup;
+ return 1;
}
- ctrl = virPCIDeviceRead16(dev, fd, pos + PCI_EXT_ACS_CTRL);
+ pci_device_cfg_read_u16(dev->device, &ctrl, pos + PCI_EXT_ACS_CTRL);
if ((ctrl & PCI_EXT_CAP_ACS_ENABLED) != PCI_EXT_CAP_ACS_ENABLED) {
VIR_DEBUG("%s %s: downstream port has ACS disabled",
dev->id, dev->name);
- ret = 1;
- goto cleanup;
+ return 1;
}
- cleanup:
- virPCIDeviceConfigClose(dev, fd);
- return ret;
+ return 0;
}
static int
@@ -3189,48 +2982,27 @@ virPCIDeviceGetVPD(virPCIDevice *dev G_GNUC_UNUSED)
int
virPCIDeviceIsPCIExpress(virPCIDevice *dev)
{
- int fd;
- int ret = -1;
-
- if ((fd = virPCIDeviceConfigOpen(dev)) < 0)
- return ret;
-
- if (virPCIDeviceInit(dev, fd) < 0)
- goto cleanup;
-
- ret = dev->is_pcie;
+ if (virPCIDeviceInit(dev) < 0)
+ return -1;
- cleanup:
- virPCIDeviceConfigClose(dev, fd);
- return ret;
+ return dev->is_pcie;
}
int
virPCIDeviceHasPCIExpressLink(virPCIDevice *dev)
{
- int fd;
- int ret = -1;
uint16_t cap, type;
-
- if ((fd = virPCIDeviceConfigOpen(dev)) < 0)
- return ret;
-
- if (virPCIDeviceInit(dev, fd) < 0)
- goto cleanup;
+ if (virPCIDeviceInit(dev) < 0)
+ return -1;
if (dev->pcie_cap_pos == 0) {
- ret = 0;
- goto cleanup;
+ return 0;
}
- cap = virPCIDeviceRead16(dev, fd, dev->pcie_cap_pos + PCI_CAP_FLAGS);
+ pci_device_cfg_read_u16(dev->device, &cap, dev->pcie_cap_pos + PCI_CAP_FLAGS);
type = (cap & PCI_EXP_FLAGS_TYPE) >> 4;
- ret = type != PCI_EXP_TYPE_ROOT_INT_EP && type != PCI_EXP_TYPE_ROOT_EC;
-
- cleanup:
- virPCIDeviceConfigClose(dev, fd);
- return ret;
+ return type != PCI_EXP_TYPE_ROOT_INT_EP && type != PCI_EXP_TYPE_ROOT_EC;
}
int
@@ -3242,53 +3014,39 @@ virPCIDeviceGetLinkCapSta(virPCIDevice *dev,
unsigned int *sta_width)
{
uint32_t t;
- int fd;
- int ret = -1;
-
- if ((fd = virPCIDeviceConfigOpen(dev)) < 0)
- return ret;
-
- if (virPCIDeviceInit(dev, fd) < 0)
- goto cleanup;
+ uint16_t t16;
+ if (virPCIDeviceInit(dev) < 0)
+ return -1;
if (!dev->pcie_cap_pos) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("pci device %1$s is not a PCI-Express device"),
dev->name);
- goto cleanup;
+ return -1;
}
- t = virPCIDeviceRead32(dev, fd, dev->pcie_cap_pos + PCI_EXP_LNKCAP);
+ pci_device_cfg_read_u32(dev->device, &t, dev->pcie_cap_pos + PCI_EXP_LNKCAP);
*cap_port = t >> 24;
*cap_speed = t & PCI_EXP_LNKCAP_SPEED;
*cap_width = (t & PCI_EXP_LNKCAP_WIDTH) >> 4;
- t = virPCIDeviceRead16(dev, fd, dev->pcie_cap_pos + PCI_EXP_LNKSTA);
+ pci_device_cfg_read_u16(dev->device, &t16, dev->pcie_cap_pos + PCI_EXP_LNKSTA);
+ t = t16;
*sta_speed = t & PCI_EXP_LNKSTA_SPEED;
*sta_width = (t & PCI_EXP_LNKSTA_WIDTH) >> 4;
- ret = 0;
-
- cleanup:
- virPCIDeviceConfigClose(dev, fd);
- return ret;
+ return 0;
}
int virPCIGetHeaderType(virPCIDevice *dev, int *hdrType)
{
- int fd;
uint8_t type;
*hdrType = -1;
- if ((fd = virPCIDeviceConfigOpen(dev)) < 0)
- return -1;
-
- type = virPCIDeviceRead8(dev, fd, PCI_HEADER_TYPE);
-
- virPCIDeviceConfigClose(dev, fd);
+ pci_device_cfg_read_u8(dev->device, &type, PCI_HEADER_TYPE);
type &= PCI_HEADER_TYPE_MASK;
if (type >= VIR_PCI_HEADER_LAST) {
diff --git a/tests/qemuhotplugtest.c b/tests/qemuhotplugtest.c
index d2a1f5acf1..fd8339bd30 100644
--- a/tests/qemuhotplugtest.c
+++ b/tests/qemuhotplugtest.c
@@ -18,12 +18,14 @@
*/
#include <config.h>
+#include "testutils.h"
+
+#ifdef __linux__
#include "qemu/qemu_alias.h"
#include "qemu/qemu_conf.h"
#include "qemu/qemu_hotplug.h"
#include "qemumonitortestutils.h"
-#include "testutils.h"
#include "testutilsqemu.h"
#include "testutilsqemuschema.h"
#include "virhostdev.h"
@@ -816,3 +818,10 @@ VIR_TEST_MAIN_PRELOAD(mymain,
VIR_TEST_MOCK("domaincaps"),
VIR_TEST_MOCK("virprocess"),
VIR_TEST_MOCK("qemuhotplug"));
+#else
+int
+main(void)
+{
+ return EXIT_AM_SKIP;
+}
+#endif
diff --git a/tests/qemumemlocktest.c b/tests/qemumemlocktest.c
index a2888732f7..5f62c80665 100644
--- a/tests/qemumemlocktest.c
+++ b/tests/qemumemlocktest.c
@@ -7,6 +7,8 @@
#include "testutils.h"
+#ifdef __linux__
+
#ifdef WITH_QEMU
# include "internal.h"
@@ -137,3 +139,10 @@ main(void)
}
#endif /* WITH_QEMU */
+#else
+int
+main(void)
+{
+ return EXIT_AM_SKIP;
+}
+#endif
diff --git a/tests/qemuxmlconftest.c b/tests/qemuxmlconftest.c
index 1f0068864a..6f84812109 100644
--- a/tests/qemuxmlconftest.c
+++ b/tests/qemuxmlconftest.c
@@ -7,6 +7,8 @@
#include "testutils.h"
+#ifdef __linux__
+
#ifdef WITH_QEMU
# include "internal.h"
@@ -3051,3 +3053,10 @@ int main(void)
}
#endif /* WITH_QEMU */
+#else
+int
+main(void)
+{
+ return EXIT_AM_SKIP;
+}
+#endif
diff --git a/tests/virpcimock.c b/tests/virpcimock.c
index 5b923c63ce..2fa12903d2 100644
--- a/tests/virpcimock.c
+++ b/tests/virpcimock.c
@@ -22,7 +22,7 @@
#include "virpcivpdpriv.h"
-#if defined(__linux__) || defined(__FreeBSD__) || defined(__APPLE__)
+#if defined(__linux__) || defined(__APPLE__)
# define VIR_MOCK_LOOKUP_MAIN
# include "virmock.h"
# include "virpci.h"
@@ -42,6 +42,10 @@ static int (*real___open_2)(const char *path, int flags);
static int (*real_close)(int fd);
static DIR * (*real_opendir)(const char *name);
static char *(*real_virFileCanonicalizePath)(const char *path);
+static int (*real_scandir)(const char *restrict dirp,
+ struct dirent ***restrict namelist,
+ typeof(int(const struct dirent *)) *filter,
+ typeof(int(const struct dirent **, const struct dirent **)) *compar);
static char *fakerootdir;
@@ -955,6 +959,7 @@ init_syms(void)
VIR_MOCK_REAL_INIT(opendir);
# endif
VIR_MOCK_REAL_INIT(virFileCanonicalizePath);
+ VIR_MOCK_REAL_INIT(scandir);
}
static void
@@ -1172,6 +1177,20 @@ virFileCanonicalizePath(const char *path)
return real_virFileCanonicalizePath(newpath);
}
+int scandir(const char *restrict dirp, struct dirent ***restrict namelist,
+ typeof(int(const struct dirent *)) *filter,
+ typeof(int(const struct dirent **, const struct dirent **)) *compar)
+{
+ g_autofree char *newpath = NULL;
+
+ init_syms();
+
+ if (getrealpath(&newpath, dirp) < 0)
+ return -1;
+
+ return real_scandir(newpath, namelist, filter, compar);
+}
+
# include "virmockstathelpers.c"
#else
--
2.47.1
3 weeks, 6 days
[PATCH 0/2] remote_daemon: Silence DBus errors
by Michal Privoznik
*** BLURB HERE ***
Michal Prívozník (2):
virgdbus: Introduce virGDBusHasSessionBus()
remote_daemon: Silence DBus errors
src/libvirt_private.syms | 1 +
src/remote/remote_daemon.c | 36 +++++++++++++++------------
src/util/virgdbus.c | 50 +++++++++++++++++++++++++++++++++++---
src/util/virgdbus.h | 3 +++
4 files changed, 70 insertions(+), 20 deletions(-)
--
2.45.3
3 weeks, 6 days
[PATCH v2 0/8] Refactor autostart for VM daemons
by Daniel P. Berrangé
These patches are a spin off from my bigger auto-shutdown series:
https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/EI...
Since they're self contained, they can be merged sooner while
I'm still working on Peter's feedback from latter parts of the
above patch series.
Daniel P. Berrangé (8):
bhyve: add 'driver' to bhyveProcessStart method
bhyve: stop opening & passing connection to bhyveProcessStart
util: add APIs for more systemd notifications
remote: notify systemd when reloading config
hypervisor: introduce helper for autostart
src: convert drivers over to use new autostart helper
hypervisor: add support for delay interval during autostart
qemu: add 'auto_start_delay' configuration parameter
src/bhyve/bhyve_driver.c | 60 ++++++++++--------------------
src/bhyve/bhyve_process.c | 5 +--
src/bhyve/bhyve_process.h | 3 +-
src/hypervisor/domain_driver.c | 55 +++++++++++++++++++++++++++
src/hypervisor/domain_driver.h | 19 ++++++++++
src/libvirt_private.syms | 7 +++-
src/libxl/libxl_driver.c | 36 +++++-------------
src/lxc/lxc_driver.c | 13 ++++---
src/lxc/lxc_process.c | 18 +--------
src/lxc/lxc_process.h | 2 +
src/qemu/libvirtd_qemu.aug | 1 +
src/qemu/qemu.conf.in | 5 +++
src/qemu/qemu_conf.c | 2 +
src/qemu/qemu_conf.h | 1 +
src/qemu/qemu_driver.c | 55 +++++++++------------------
src/qemu/test_libvirtd_qemu.aug.in | 1 +
src/remote/libvirtd.service.in | 2 +-
src/remote/remote_daemon.c | 2 +
src/rpc/virnetdaemon.c | 2 +-
src/util/virsystemd.c | 44 ++++++++++++++++++++--
src/util/virsystemd.h | 6 ++-
src/virtd.service.in | 2 +-
22 files changed, 202 insertions(+), 139 deletions(-)
--
2.47.1
4 weeks
Re: [PATCH 2/2] nbd/server: Allow users to adjust handshake limit in
QMP
by Eric Blake
On Thu, Feb 06, 2025 at 10:20:09AM +0300, Vladimir Sementsov-Ogievskiy wrote:
> > ---
> > qapi/block-export.json | 10 ++++++++++
> > include/block/nbd.h | 6 +++---
>
> [..]
>
> > @@ -52,6 +57,10 @@
> > #
> > # @addr: Address on which to listen.
> > #
> > +# @handshake-max-secs: Time limit, in seconds, at which a client that
> > +# has not completed the negotiation handshake will be disconnected,
> > +# or 0 for no limit (since 10.0; default: 10).
> > +#
>
> Hmm. [not about the series], shouldn't we finally deprecate older interface?
By older interface, you are asking about the QMP command
'nbd-server-start' as compared to struct NbdServerOptions. But the
struct is not directly present in any QMP commands; rather, it only
appears to be used by qemu-storage-daemon as one of its command line
options that needs to set up an NBD server with a JSON-like syntax
that has less nesting than QMP nbd-server-start. blockdev-nbd.c has
two functions [nbd_server_start_options(NbdServerOPtions *arg...) and
qmp_nbd_server_start(args...)] that both unpack their slightly
different forms and pass them as parameters to nbd_server_start() that
is then agnostic to whether the older QMP command or newer q-s-d CLI
option was specified.
It looks like libvirt is still using QMP nbd-server-start. If we were
to start the deprecation process for qemu 10.0, what would the new
command look like? What would everyone be required to use by qemu
10.2?
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
4 weeks
[PATCH 0/5] qemu: migration: Reactivate block nodes after migration if VM is left paused
by Peter Krempa
Patch 1/5 is a bugfix for the qemu-replies-tool.
For justification of 2-5/5 see 5/5.
Peter Krempa (5):
qemu-replies-tool: Dump also query string for optional members
qemucapabilitiestest: Update 'caps_10.0.0_x86_64' to
'v9.2.0-1636-gffaf7f0376'
qemu: capabilies: Introduce QEMU_CAPS_BLOCKDEV_SET_ACTIVE
qemu: monitor: Add monitor backend for 'blockdev-set-active'
qemu: migration: Reactivate block nodes after migration if VM is left
paused
scripts/qemu-replies-tool.py | 4 +
src/qemu/qemu_capabilities.c | 2 +
src/qemu/qemu_capabilities.h | 1 +
src/qemu/qemu_migration.c | 46 +-
src/qemu/qemu_monitor.c | 21 +
src/qemu/qemu_monitor.h | 5 +
src/qemu/qemu_monitor_json.c | 21 +
src/qemu/qemu_monitor_json.h | 5 +
.../caps_10.0.0_x86_64.replies | 3918 +++++++++--------
.../caps_10.0.0_x86_64.xml | 3 +-
tests/qemumonitorjsontest.c | 31 +
11 files changed, 2113 insertions(+), 1944 deletions(-)
--
2.48.1
4 weeks
[PATCH v5 00/18] *** qemu: block: Support block disk along with throttle filters ***
by Harikumar R
*** BLURB HERE ***
Chun Feng Wu (17):
schema: Add new domain elements to support multiple throttle groups
schema: Add new domain elements to support multiple throttle filters
config: Introduce ThrottleGroup and corresponding XML parsing
config: Introduce ThrottleFilter and corresponding XML parsing
qemu: monitor: Add support for ThrottleGroup operations
tests: Test qemuMonitorJSONGetThrottleGroup and
qemuMonitorJSONUpdateThrottleGroup
remote: New APIs for ThrottleGroup lifecycle management
qemu: Refactor qemuDomainSetBlockIoTune to extract common methods
qemu: Implement qemu driver for throttle API
qemu: helper: throttle filter nodename and preparation processing
qemu: block: Support block disk along with throttle filters
config: validate: Verify iotune, throttle group and filter
qemuxmlconftest: Add 'throttlefilter' tests
test_driver: Test throttle group lifecycle APIs
virsh: Refactor iotune options for re-use
virsh: Add support for throttle group operations
virsh: Add option "throttle-groups" to "attach_disk"
Harikumar Rajkumar (1):
tests: Test qemuxmlactivetestThrottleGroup
docs/formatdomain.rst | 47 ++
docs/manpages/virsh.rst | 135 +++-
include/libvirt/libvirt-domain.h | 21 +
src/conf/domain_conf.c | 398 ++++++++++
src/conf/domain_conf.h | 45 ++
src/conf/domain_validate.c | 119 ++-
src/conf/schemas/domaincommon.rng | 293 ++++----
src/conf/virconftypes.h | 4 +
src/driver-hypervisor.h | 22 +
src/libvirt-domain.c | 174 +++++
src/libvirt_private.syms | 8 +
src/libvirt_public.syms | 7 +
src/qemu/qemu_block.c | 136 ++++
src/qemu/qemu_block.h | 49 ++
src/qemu/qemu_command.c | 180 +++++
src/qemu/qemu_command.h | 6 +
src/qemu/qemu_domain.c | 73 +-
src/qemu/qemu_driver.c | 619 +++++++++++++---
src/qemu/qemu_hotplug.c | 33 +
src/qemu/qemu_monitor.c | 34 +
src/qemu/qemu_monitor.h | 14 +
src/qemu/qemu_monitor_json.c | 134 ++++
src/qemu/qemu_monitor_json.h | 14 +
src/remote/remote_daemon_dispatch.c | 44 ++
src/remote/remote_driver.c | 40 ++
src/remote/remote_protocol.x | 48 +-
src/remote_protocol-structs | 28 +
src/test/test_driver.c | 452 ++++++++----
tests/qemumonitorjsontest.c | 86 +++
.../throttlefilter-in.xml | 392 ++++++++++
.../throttlefilter-out.xml | 393 ++++++++++
tests/qemuxmlactivetest.c | 1 +
.../throttlefilter-invalid.x86_64-latest.err | 1 +
.../throttlefilter-invalid.xml | 89 +++
.../throttlefilter.x86_64-latest.args | 55 ++
.../throttlefilter.x86_64-latest.xml | 105 +++
tests/qemuxmlconfdata/throttlefilter.xml | 95 +++
tests/qemuxmlconftest.c | 2 +
tools/virsh-completer-domain.c | 82 +++
tools/virsh-completer-domain.h | 16 +
tools/virsh-domain.c | 680 ++++++++++++++----
41 files changed, 4649 insertions(+), 525 deletions(-)
create mode 100644 tests/qemustatusxml2xmldata/throttlefilter-in.xml
create mode 100644 tests/qemustatusxml2xmldata/throttlefilter-out.xml
create mode 100644 tests/qemuxmlconfdata/throttlefilter-invalid.x86_64-latest.err
create mode 100644 tests/qemuxmlconfdata/throttlefilter-invalid.xml
create mode 100644 tests/qemuxmlconfdata/throttlefilter.x86_64-latest.args
create mode 100644 tests/qemuxmlconfdata/throttlefilter.x86_64-latest.xml
create mode 100644 tests/qemuxmlconfdata/throttlefilter.xml
--
2.39.5 (Apple Git-154)
4 weeks
[PATCH] node_device: Do not lock the driver state needlessly
by Martin Kletzander
When processing the PCI devices we can only read the configs for each of
them if running as privileged. That information is saved in the driver
state as a boolean introduced in commit 643c74abff01. However since
that version it is only written to once during nodeStateInitialize() and
only read from that point (apart from some commits around v3.9.0 release
when it was not even set, but that was fixed before v3.10.0). And it is
only read once, just to store that boolean in a temporary variable which
is also used in only one condition.
Rewrite this without locking and save few lines of code.
Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com>
---
src/node_device/node_device_udev.c | 7 +------
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c
index 7b4ff80f8fb4..be5ee0108c85 100644
--- a/src/node_device/node_device_udev.c
+++ b/src/node_device/node_device_udev.c
@@ -430,7 +430,6 @@ udevProcessPCI(virNodeDeviceDriverState *driver_state,
g_autofree char *linkpath = NULL;
int ret = -1;
char *p;
- bool privileged = false;
linkpath = g_strdup_printf("%s/config", udev_device_get_syspath(device));
if (virFileWaitForExists(linkpath, 10, 100) < 0) {
@@ -440,10 +439,6 @@ udevProcessPCI(virNodeDeviceDriverState *driver_state,
goto cleanup;
}
- VIR_WITH_MUTEX_LOCK_GUARD(&driver_state->lock) {
- privileged = driver_state->privileged;
- }
-
pci_dev->klass = -1;
if (udevGetIntProperty(device, "PCI_CLASS", &pci_dev->klass, 16) < 0)
goto cleanup;
@@ -491,7 +486,7 @@ udevProcessPCI(virNodeDeviceDriverState *driver_state,
goto cleanup;
/* We need to be root to read PCI device configs */
- if (privileged) {
+ if (driver_state->privileged) {
if (virPCIGetHeaderType(pciDev, &pci_dev->hdrType) < 0)
goto cleanup;
--
2.48.1
4 weeks
[libvirt PATCH] qemu: fix qemuDomainSaveImageDefineXML
by Pavel Hrdina
The commit in question made an incorrect change that resulted in getting
O_RDONLY FD instead of O_RDWR preventing any writes to happen with the
following error:
virQEMUSaveDataWrite:176 : failed to write header to domain save file '/path/to/save.img': Bad file descriptor
Fixes: 517248e2394476a3105ff5866b0b718fc6583073
Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com>
---
src/qemu/qemu_driver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
index 78bfaa5b3a..a35abf2747 100644
--- a/src/qemu/qemu_driver.c
+++ b/src/qemu/qemu_driver.c
@@ -5949,7 +5949,7 @@ qemuDomainSaveImageDefineXML(virConnectPtr conn, const char *path,
if (qemuSaveImageGetMetadata(driver, NULL, path, &def, &data) < 0)
goto cleanup;
- fd = qemuSaveImageOpen(driver, path, 0, NULL, false);
+ fd = qemuSaveImageOpen(driver, path, false, NULL, true);
if (fd < 0)
goto cleanup;
--
2.48.1
4 weeks
