Drives defined with -drive if=ide get get created along with the IDE
controller, inside machine->init(). That's before cmos_init().
Drives defined with -device get created during generic device init.
That's after cmos_init(). Because of that, CMOS has no information on
them (type, geometry, translation). Older versions of Windows such as
XP reportedly choke on that.
Split off the part of CMOS initialization that needs to know about
-device devices, and turn it into a reset handler, so it runs after
device creation.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
BlockDriverState member removable controls whether virtual media
change (monitor commands change, eject) is allowed. It is set when
the "type hint" is BDRV_TYPE_CDROM or BDRV_TYPE_FLOPPY.
The type hint is only set by drive_init(). It sets BDRV_TYPE_FLOPPY
for if=floppy. It sets BDRV_TYPE_CDROM for media=cdrom and if=ide,
scsi, xen, or none.
if=ide and if=scsi work, because the type hint makes it a CD-ROM.
if=xen likewise, I think.
For the same reason, if=none works when it's used by ide-drive or
scsi-disk. For other guest devices, there are problems:
* fdc: you can't change virtual media
$ qemu [...] -drive if=none,id=foo,... -global isa-fdc.driveA=foo
QEMU 0.12.50 monitor - type 'help' for more information
(qemu) eject foo
Device 'foo' is not removable
unless you add media=cdrom, but that makes it readonly.
* virtio: if you add media=cdrom, you can change virtual media. If
you eject, the guest gets I/O errors. If you change, the guest sees
the drive's contents suddenly change.
* scsi-generic: if you add media=cdrom, you can change virtual media.
I didn't test what that does to the guest or the physical device,
but it can't be pretty.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
savevm.c keeps a pointer to the snapshot block device. If you manage
to get that device deleted, the pointer dangles, and the next snapshot
operation will crash & burn. Unplugging a guest device that uses it
does the trick:
$ MALLOC_PERTURB_=234 qemu-system-x86_64 [...]
QEMU 0.12.50 monitor - type 'help' for more information
(qemu) info snapshots
No available block device supports snapshots
(qemu) drive_add auto if=none,file=tmp.qcow2
OK
(qemu) device_add usb-storage,id=foo,drive=none1
(qemu) info snapshots
Snapshot devices: none1
Snapshot list (from none1):
ID TAG VM SIZE DATE VM CLOCK
(qemu) device_del foo
(qemu) info snapshots
Snapshot devices:
Segmentation fault (core dumped)
Move management of that pointer to block.c, and zap it when the device
it points becomes unusable.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Kevin Wolf [Wed, 30 Jun 2010 15:43:40 +0000 (17:43 +0200)]
blkdebug: Initialize state as 1
state = 0 in rules means that the rule is valid for any state. Therefore it's
impossible to have a rule that works only in the initial state. This changes
the initial state from 0 to 1 to make this possible.
block: Catch attempt to attach multiple devices to a blockdev
For instance, -device scsi-disk,drive=foo -device scsi-disk,drive=foo
happily creates two SCSI disks connected to the same block device.
It's all downhill from there.
Device usb-storage deliberately attaches twice to the same blockdev,
which fails with the fix in place. Detach before the second attach
there.
Also catch attempt to delete while a guest device model is attached.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Make the property point to BlockDriverState, cutting out the DriveInfo
middleman. This prepares the ground for block devices that don't have
a DriveInfo.
Currently all user-defined ones have a DriveInfo, because the only way
to define one is -drive & friends (they go through drive_init()).
DriveInfo is closely tied to -drive, and like -drive, it mixes
information about host and guest part of the block device. I'm
working towards a new way to define block devices, with clean
host/guest separation, and I need to get DriveInfo out of the way for
that.
Fortunately, the device models are perfectly happy with
BlockDriverState, except for two places: ide_drive_initfn() and
scsi_disk_initfn() need to check the DriveInfo for a serial number set
with legacy -drive serial=... Use drive_get_by_blockdev() there.
Device model code should now use DriveInfo only when explicitly
dealing with drives defined the old way, i.e. without -device.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
We automatically delete blockdev host parts on unplug of the guest
device. Too much magic, but we can't change that now.
The delete happens early in the guest device teardown, before the
connection to the host part is severed. Thus, the guest part's
pointer to the host part dangles for a brief time. No actual harm
comes from this, but we'll catch such dangling pointers a few commits
down the road. Clean up the dangling pointers by delaying the
automatic deletion until the guest part's pointer is gone.
Device usb-storage deliberately makes two qdev properties refer to the
same drive, because it automatically creates a second device. Again,
too much magic we can't change now. Multiple references worked okay
before, but now free_drive() dies for the second one. Zap the extra
reference.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Ryan Harper [Mon, 28 Jun 2010 14:38:33 +0000 (09:38 -0500)]
Don't reset bs->is_temporary in bdrv_open_common
To fix https://bugs.launchpad.net/qemu/+bug/597402 where qemu fails to
call unlink() on temporary snapshots due to bs->is_temporary getting clobbered
in bdrv_open_common() after being set in bdrv_open() which calls the former.
We don't need to initialize bs->is_temporary in bdrv_open_common().
Signed-off-by: Ryan Harper <ryanh@us.ibm.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
ide: Make it explicit that ide_create_drive() can't fail
All callers of ide_create_drive() ignore its value. Currently
harmless, because it fails only when qdev_init() fails, which fails
only when ide_drive_initfn() fails, which never fails.
Brittle. Change it to die instead of silently ignoring failure.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
block: allow filenames with colons again for host devices
Before the raw/file split we used to allow filenames with colons for host
device only. While this was more by accident than by design people rely
on it, so we need to bring it back.
So move the host device probing to be before the protocol detection
again.
Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Andi Kleen [Sat, 26 Jun 2010 22:06:11 +0000 (00:06 +0200)]
Add more boundary checking to sse3/4 parsing
ssse3 uses tables with only two entries per op, but it is indexed
with b1 which can contain variables upto 3. This happens when ssse3
or sse4 are used with REP* prefixes.
Kevin Wolf [Fri, 18 Jun 2010 16:27:03 +0000 (18:27 +0200)]
qdev-properties: Fix (u)intXX parsers
scanf calls must not use PRI constants, they have probably the wrong size and
corrupt memory. We could replace them by SCN ones, but strtol is simpler than
scanf here anyway. While at it, also fix the parsers to reject garbage after
the number ("4096xyz" was accepted before).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Richard Henderson <rth@redhat.com> Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Amit Shah [Tue, 15 Jun 2010 08:00:39 +0000 (13:30 +0530)]
net: Fix VM start with '-net none'
Commit 50e32ea8f31035877decc10f1075aa0e619e09cb changed the behaviour
for the return type of net_client_init() when a nic type with no init
method was specified. 'none' is one such nic type. Instead of returning
0, which gets interpreted as an index into the nd_table[] array, we
switched to returning -1, which signifies an error as well.
That broke VM start with '-net none'. Testing was only done with the
monitor command 'pci_add', which doesn't fail.
The correct fix would still be to return 0+ values from
net_client_init() only when the return value can be used as an index to
refer to an entry in nd_table[]. With the current code, callers can
erroneously poke into nd_table[0] when -net nic is used, which can lead
to badness.
However, this commit just returns to the previous behaviour before the
offending commit.
Signed-off-by: Amit Shah <amit.shah@redhat.com> Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Jan Kiszka [Wed, 2 Jun 2010 06:49:14 +0000 (08:49 +0200)]
x86: svm: Always clear event_inj on vmexit
We currently only clear SVM_EVTINJ_VALID after successful interrupt
delivery. This apparently does not match real hardware which clears the
whole event_inj field on every vmexit, including unsuccessful interrupt
delivery.
Reported-by: Erik van der Kouwe <vdkouwe@cs.vu.nl> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
The commit 8e65b7c04965c8355e4ce43211582b6b83054e3d introduced
expire_time of UHCIState. But expire_time is not in vmstate, the
second uhci_frame_timer will not be fired immediately after loadvm.
Jun Koi [Thu, 6 May 2010 05:36:59 +0000 (14:36 +0900)]
A bit optimization for tlb_set_page()
This patch avoids handling write watchpoints on read-only memory access.
It also breaks the searching loop for watchpoint once the setup for
handling watchpoint later is done.
Signed-off-by: Jun Koi <junkoi2004@gmail.com> Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Paolo Bonzini [Mon, 14 Jun 2010 17:11:54 +0000 (19:11 +0200)]
lsi53c895a: fix Phase Mismatch Jump
lsi_bad_phase has a bug in the choice of pmjad1/pmjad2. This does
not matter with Linux guests because it uses just one routine for
both, but it breaks Windows 64-bit guests. This is the text
from the spec:
"[The PMJCTL] bit controls which decision mechanism is used
when jumping on phase mismatch. When this bit is cleared the
LSI53C895A will use Phase Mismatch Jump Address 1 (PMJAD1) when
the WSR bit is cleared and Phase Mismatch Jump Address 2 (PMJAD2)
when the WSR bit is set. When this bit is set the LSI53C895A will
use jump address one (PMJAD1) on data out (data out, command,
message out) transfers and jump address two (PMJAD2) on data in
(data in, status, message in) transfers."
Which means:
CCNTL0.PMJCTL
0 SCNTL2.WSR = 0 PMJAD1
0 SCNTL2.WSR = 1 PMJAD2
1 out PMJAD1
1 in PMJAD2
In qemu, what you get instead is:
CCNTL0.PMJCTL
0 out PMJAD1
0 in PMJAD2 <<<<<
1 out PMJAD1
1 in PMJAD1 <<<<<
Considering that qemu always has SCNTL2.WSR cleared, the two marked cases
(corresponding to phase mismatch on input) are always jumping to the
wrong PMJAD register. The patch implements the correct semantics.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Andre Przywara [Wed, 2 Jun 2010 09:57:47 +0000 (11:57 +0200)]
fix CPUID vendor override
the meaning of vendor_override is actually the opposite of how it
is currently used :-(
Fix it to allow KVM to export the non-native CPUID vendor if
explicitly requested by the user.
The intended behavior is:
With TCG:
- always inject the configured vendor (either hard-coded, in config
files or via ",vendor=" commandline)
With KVM:
- by default inject the host's vendor
- if the user specifies ",vendor=" on the commandline, use this
instead of the host's vendor
- all pre-configured vendors (hard-coded, config file) are ignored
Signed-off-by: Andre Przywara <andre.przywara@amd.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Jan Kiszka [Wed, 19 May 2010 22:28:45 +0000 (00:28 +0200)]
kvm: Switch kvm_update_guest_debug to run_on_cpu
Guest debugging under KVM is currently broken once io-threads are
enabled. Easily fixable by switching the fake on_vcpu to the real
run_on_cpu implementation.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by: Avi Kivity <avi@redhat.com>
Stefan Weil [Thu, 24 Jun 2010 20:41:33 +0000 (22:41 +0200)]
win32: Add define for missing EPROTONOSUPPORT
mingw32 does not define EPROTONOSUPPORT (which is used by
migration.c and maybe future patches), so add a
definition which uses a supported errno value.
Signed-off-by: Stefan Weil <weil@mail.berlios.de> Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Artyom Tarasenko [Mon, 21 Jun 2010 18:23:21 +0000 (20:23 +0200)]
mask all interrupts when MASTER_DISABLE is set
The MASTER_DISABLE bit (aka mask-all) masks all the interrupts.
According to Sun-4M System Architecture
"The level–15 interrupt sources [...] are maskable with the Interrupt Target
Mask Register. While these interrupts are considered ’non–maskable’ within
the SPARC IU, a mask capability is provided to allow the boot firmware
to establish a basic environment before receiving any level–15 interrupts,
which are non–maskable within SPARC. A mask–all bit is provided to allow
disabling of all external interrupts during change of the CIT."
Signed-off-by: Artyom Tarasenko <atar4qemu@gmail.com> Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Blue Swirl [Sun, 27 Jun 2010 16:04:31 +0000 (16:04 +0000)]
Remove useless device dependency of HAS_AUDIO
System architecture dictates whether HAS_AUDIO is defined. It's then
useless to check for HAS_AUDIO in files which are only used on those
architectures which always have audio.
Alex Williamson [Thu, 17 Jun 2010 15:15:02 +0000 (09:15 -0600)]
virtio-pci: fix bus master bug setting on load
The comment suggests we're checking for the driver in the ready
state and bus master disabled, but the code is checking that it's
not in the ready state.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Found-by: Amit Shah <amit.shah@redhat.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Mapped mode stores extended attributes in the user space of the extended
attributes. Given that the user space extended attributes are available
to regular files only, special files are created as regular files on the
fileserver and appropriate mode bits are added to the extended attributes.
This method presents all special files and symlinks as regular files on the
fileserver while they are represented as special files on the guest mount.
virtio-9p: Security model for symlink and readlink
Mapped mode stores extended attributes in the user space of the extended
attributes. Given that the user space extended attributes are available
to regular files only, special files are created as regular files on the
fileserver and appropriate mode bits are added to the extended attributes.
This method presents all special files and symlinks as regular files on the
fileserver while they are represented as special files on the guest mount.
Implemntation of symlink in mapped security model:
A regular file is created and the link target is written to it.
readlink() reads it back from the file.
In the mapped security model, VirtFS server intercepts and maps
the file object create and get/set attribute requests. Files on the fileserver
will be created with VirtFS servers (QEMU) user credentials and the
client-users credentials are stored in extended attributes. On the request
to get attributes, server extracts the client-users credentials
from extended attributes and sends them to the client.
On Host/Fileserver:
-rw-------. 2 virfsuid virtfsgid 0 2010-05-11 09:19 afile
On Guest/Client:
-rw-r--r-- 2 guestuser guestuser 0 2010-05-11 12:19 afile
Signed-off-by: Venkateswararao Jujjuri <jvrao@linux.vnet.ibm.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
mapped model changes the owner in the extended attributes.
passthrough model does the change through lchown() as the
server don't need to follow the link and client will send the
actual filesystem object.
Signed-off-by: Venkateswararao Jujjuri <jvrao@linux.vnet.ibm.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
virtio-9p: Make infrastructure for the new security model.
This patch adds required infrastructure for the new security model.
- A new configure option for attr/xattr.
- if CONFIG_VIRTFS will be defined if both CONFIG_LINUX and CONFIG_ATTR defined.
- Defines routines related to both security models.
Signed-off-by: Venkateswararao Jujjuri <jvrao@linux.vnet.ibm.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
In the case of mapped security model, files are created with QEMU user
credentials and the client-user's credentials are saved in extended attributes.
Whereas in the case of passthrough security model, files on the
filesystem are directly created with client-user's credentials.
Signed-off-by: Venkateswararao Jujjuri <jvrao@linux.vnet.ibm.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Michael Tokarev [Wed, 2 Jun 2010 17:33:01 +0000 (14:33 -0300)]
give some useful error messages when tap open
In net/tap-linux.c, when manipulation of /dev/net/tun fails, it prints
(with fprintf) something like this:
warning: could not open /dev/net/tun: no virtual network emulation
this has 2 issues:
1) it is not a warning really, it's a fatal error (kvm exits after
that),
2) there's no indication as of what's actually wrong: printing errno there
is helpful.
The patch below removes the "warning" prefix, uses %m (since it's linux,
%m is available as format modifier), and changes fprintf() to %qemu_error().
Now it prints something like this instead:
could not configure /dev/net/tun: Device or resource busy
(there are 2 messages like that in the same function)
This fixes Debian bug #578154, see
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578154
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
projects / qemu-xen.git / log