KVM_GET_SUPPORTED_CPUID has been known to fail to return -E2BIG
when it runs out of entries. Detect this by always trying again
with a bigger table if the ioctl() fills the table.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Remove cpu features that are not supported by kvm from the cpuid features
reported to the guest.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
QEMU allows adding or removing cpu features by using the syntax '-cpu +feature'
or '-cpu -feature'. Some cpuid features cause more than one bit to be set or
cleared; but QEMU stops after just one bit has been modified, causing the
feature bits to be inconsistent.
Fix by allowing all feature bits corresponding to a given name to be set.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
To be used later.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
kvm does not support all cpu features; add support for dunamically querying
the supported feature set.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Right now, if you try e.g. '-smp 2' you just get 'failed to
initialize KVM'.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
When qemu is run under valgrind, valgrind shows the following output
on exit:
==3648== 1 errors in context 2 of 2:
==3648== Syscall param timer_create(evp) points to uninitialised byte(s)
==3648== at 0x54E936A: timer_create (in /lib/librt-2.9.so)
==3648== by 0x405DCF: dynticks_start_timer (vl.c:1549)
==3648== by 0x40A966: main (vl.c:1726)
==3648== Address 0x7fefffb34 is on thread 1's stack
==3648== Uninitialised value was created by a stack allocation
==3648== at 0x405D60: dynticks_start_timer (vl.c:1534)
This patch is a simple fix to remove this potential problem.
Signed-off-by: Jean-Christophe DUBOIS <jcd@tribudubois.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This fixes a SIGSEGV error on qemu exit.
Here is the valgrind output related to this error
==3648== Process terminating with default action of signal 11 (SIGSEGV)
==3648== Access not within mapped region at address 0x8
==3648== at 0x40636B: host_alarm_handler (vl.c:1345)
==3648== by 0x52D807F: (within /lib/libpthread-2.9.so)
==3648== by 0x5C0A12E: tcsetattr (in /lib/libc-2.9.so)
==3648== by 0x4DD601: term_exit (qemu-char.c:700)
==3648== by 0x5B636EC: exit (in /lib/libc-2.9.so)
==3648== by 0x5B4B5AC: (below main) (in /lib/libc-2.9.so)
This simple fix check for a valid pointer as host_alarm_handler is
also called after alarm_timer is released in the exit path.
Signed-off-by: Jean-Christophe DUBOIS <jcd@tribudubois.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
initrd must be kept on the memory area below 4g. By not doing this,
we're seeing guests break while using -initrd and values of -mem
superior to 4096.
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The migration code is non-blocking, designed for live migration.
Practically migrate_fd_put_buffer busy-loops trying to write, as
on many machines EWOULDBLOCK==EAGAIN (look in include/asm-generic/errno.h).
Signed-off-by: Uri Lublin <uril@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This text is no longer accurate. After the patch is applied, the
generated version at http://www.nongnu.org/qemu/qemu-doc.html should be
regenerated.
This patch is also a candidate for the stable branch. (The URL above is
probably generated from the stable branch anyways, so maybe it goes
without saying.)
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The block layer may signal an immediate error on an asynchronous request
by returning NULL. The DMA API did not handle this correctly, returning
an AIO request which would never complete (and which would crash if
cancelled).
Fix by detecting the failure and propagating it.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6893 c046a42c-6fe2-441c-8c8c-71466251a162
If a bounced vectored aio fails immediately (the inner aio submission
returning NULL) then the bounce handler erronously returns an aio
request which will never be completed (and which crashes when cancelled).
Fix by detecting that the inner request has failed and propagating the
error.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6892 c046a42c-6fe2-441c-8c8c-71466251a162
Use the dedicated dma aiocb to store intermediate state for dma block
transactions.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6874 c046a42c-6fe2-441c-8c8c-71466251a162
Now that we have a dedicated acb pool for vector translation acbs, we can
store the vector translation state in the acbs instead of in an external
structure.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6873 c046a42c-6fe2-441c-8c8c-71466251a162
The linux loader is just an option rom like any other, just with
some special requirements. Right now, our option rom resetting
mechanism is not being applied to it. As a result, users using
-kernel will not be able to successfully reboot their machines
This patch fixes it by saving all the data we generated in
the load_linux() function, to be used later by the option rom
resetting mechanism.
This also includes Mark's fix for -kernel
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This reverts commit 2da1e398641d9fccf683645c808dee0d088f84cf.
This fix on the stable branch:
commit 2da1e398641d9fccf683645c808dee0d088f84cf
Author: Glauber Costa <glommer@redhat.com>
Date: Fri May 8 02:22:13 2009 -0300
reset state for load_linux
Caused -kernel to break.
The problem is that we're passing the ROM's ram_addr_t to
load_linux() rather than its target_phys_addr_t. We also
need to register the memory before trying to write to
it.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This patch makes the range checks for block requests more strict: It fixes a
potential integer overflow and checks for negative offsets. Also, it adds the
check for compressed writes.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
pci_register_device already mallocs the pci config space buffer filled
with zeroes.
Doing this again breaks some default config space writes like
setting the subsystem vendor id and subsystem device id.
Signed-off-by: Amit Shah <amit.shah@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
When deleting an fd event there is a chance the object doesn't get
deleted, but only ->deleted set positive and deleted somewhere later.
Now, if we create a handler for the fd again before the actual
deletion occurs, we end up writing data into an object that has
->deleted set, which is obviously wrong.
I see two ways to fix this:
1. Don't return ->deleted objects in the search
2. Unset ->deleted in the search
This patch implements 1. which feels safer to do. It fixes AIO issues
I've seen with curl, as libcurl unsets fd event listeners pretty
frequently.
Signed-off-by: Alexander Graf <alex@csgraf.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The linux loader is just an option rom like any other, just with
some special requirements. Right now, our option rom resetting
mechanism is not being applied to it. As a result, users using
-kernel will not be able to successfully reboot their machines
This patch fixes it by saving all the data we generated in
the load_linux() function, to be used later by the option rom
resetting mechanism.
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Currently, boot options are not preserved across a system reset.
option roms can modify themselves, or can for instance restore the real
int 0x19 vector after they tried to boot from it.
To properly do that, we need a reset handler registered to deal with option
roms. This patch is based on current version on qemu-kvm.git
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Need to drop QCOW_OFLAG_COPIED from a cluster pointer before freeing it.
Add an explanation how thing meant to work.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Move the dma helpers to a private aio pool, and implement a cancellation
method for them. Should prevent issues when cancelling I/O while dma is
in progress.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6872 c046a42c-6fe2-441c-8c8c-71466251a162
This allows us to remove a hack in the vectored aio cancellation code.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6871 c046a42c-6fe2-441c-8c8c-71466251a162
Move the AIOCB allocation code to use a dedicate structure, AIOPool. AIOCB
specific information, such as the AIOCB size and cancellation routine, is
moved into the pool.
At present, there is exactly one pool per block format driver, maintaining
the status quo.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6870 c046a42c-6fe2-441c-8c8c-71466251a162
Trivial build warning/fixes when the local DEBUG define is enabled.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The pci_register_device() call in PCI nic initialization routines can
fail. Handle this failure and propagate a meaningful error message to
the user instead of generating a SEGV.
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
According to PnP specification, Appendix B, Option ROMs
that support DDIM (device driver initialization model) should
have their memory space writeable.
KVM deviates from us here, by removing the IO_MEM_ROM flag,
to allow for PCI option ROMs (they require DDIM). However,
there's absolutely no reason we can't do the same.
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Avi Kivity wrote:
> Suggest wrapping in a function and hiding it deep inside kvm-all.c.
>
Done in v2:
---------->
If the KVM MMU is asynchronous (kernel does not support MMU_NOTIFIER),
we have to avoid COW for the guest memory. Otherwise we risk serious
breakage when guest pages change there physical locations due to COW
after fork. Seen when forking smbd during runtime via -smb.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The algorithm from the VHD specification for CHS calculation silently limits
images to 127 GB which may confuse a user who requested a larger image. Better
output an error message and abort.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/branches/stable_0_10@7183 c046a42c-6fe2-441c-8c8c-71466251a162
While Intel's spec is not that clear here, latest changes to Linux' HPET
code (commit c23e253e67c9d8a91a0ffa33c1f571a17f0a2403, "x86: hpet: stop
HPET_COUNTER when programming periodic mode") strongly suggest that
HPET_TN_SETVAL rather means: Set _both_ the comparator value and
register.
With this patch applied, I'm again able to boot 2.6.30-rc kernels as
they no longer panic like this (which was due to the comparator
register remaining 0):
ENABLING IO-APIC IRQs
..TIMER: vector=0x30 apic1=0 pin1=2 apic2=-1 pin2=-1
..MP-BIOS bug: 8254 timer not connected to IO-APIC
...trying to set up timer (IRQ0) through the 8259A ...
..... (found apic 0 pin 2) ...
....... failed.
...trying to set up timer as Virtual Wire IRQ...
..... failed.
...trying to set up timer as ExtINT IRQ...
..... failed :(.
Kernel panic - not syncing: IO-APIC + timer doesn't work! [...]
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/branches/stable_0_10@7171 c046a42c-6fe2-441c-8c8c-71466251a162
Fix (more or less) spurious guest boot failures due to corrupted cpuid
states. The reason was insufficient initialization of cpuid entries
before passing them to the kernel.
At this chance also fix improper entry pointer progression and simplify
the code a bit.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/branches/stable_0_10@7170 c046a42c-6fe2-441c-8c8c-71466251a162
This patch fixes a qcow2 corruption bug introduced in SVN Rev 5861. L2 tables
are big endian, so entries must be converted before being passed to functions.
This bug is easy to trigger. The following script will create and destroy a
qcow2 image (the header is gone after three loop iterations):
#!/bin/bash
qemu-img create -f qcow2 test.qcow 1M
for i in $(seq 1 10); do
qemu-system-x86_64 -hda test.qcow -monitor stdio > /dev/null 2>&1 <<EOF
savevm test-$i
quit
EOF
done
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/branches/stable_0_10@7166 c046a42c-6fe2-441c-8c8c-71466251a162
It's allocated using qemu_mallocz(), so ...
The name and model strings are strdup() allocated, so free()
is still appropriate for them.
Reported-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/branches/stable_0_10@7161 c046a42c-6fe2-441c-8c8c-71466251a162
