pc: Add a comment explaining why pc_compat_2_4() doesn't exist
pc_compat_2_4() doesn't exist, and we shouldn't create one. Add a
comment explaining why the function doesn't exist and why pc_compat_*()
functions are deprecated.
This inserts a read and write protected page between RAM and QEMU
memory, for file-backend RAM.
This makes it harder to exploit QEMU bugs resulting from buffer
overflows in devices using variants of cpu_physical_memory_map,
dma_memory_map etc.
This inserts a read and write protected page between RAM and QEMU
memory. This makes it harder to exploit QEMU bugs resulting from buffer
overflows in devices using variants of cpu_physical_memory_map,
dma_memory_map etc.
Jason Wang [Fri, 25 Sep 2015 05:21:30 +0000 (13:21 +0800)]
virtio-net: correctly drop truncated packets
When packet is truncated during receiving, we drop the packets but
neither discard the descriptor nor add and signal used
descriptor. This will lead several issues:
- sg mappings are leaked
- rx will be stalled if a lots of packets were truncated
In order to be consistent with vhost, fix by discarding the descriptor
in this case.
Jason Wang [Fri, 25 Sep 2015 05:21:29 +0000 (13:21 +0800)]
virtio: introduce virtqueue_discard()
This patch introduces virtqueue_discard() to discard a descriptor and
unmap the sgs. This will be used by the patch that will discard
descriptor when packet is truncated.
# gpg: Signature made Wed 30 Sep 2015 09:24:02 BST using RSA key ID 5872D723
# gpg: Good signature from "Juan Quintela <[email protected]>"
# gpg: aka "Juan Quintela <[email protected]>"
* remotes/juanquintela/tags/migration/20150930:
migration: Disambiguate MAX_THROTTLE
qmp/hmp: Add throttle ratio to query-migrate and info migrate
migration: Dynamic cpu throttling for auto-converge
migration: Parameters for auto-converge cpu throttling
cpu: Provide vcpu throttling interface
migration: yet more possible state transitions
Jason J. Herne [Tue, 8 Sep 2015 17:12:37 +0000 (13:12 -0400)]
migration: Disambiguate MAX_THROTTLE
Migration has a define for MAX_THROTTLE. Update comment to clarify that this is
used for throttling transfer speed. Hopefully this will prevent it from being
confused with a guest cpu throttling entity.
Jason J. Herne [Tue, 8 Sep 2015 17:12:35 +0000 (13:12 -0400)]
migration: Dynamic cpu throttling for auto-converge
Remove traditional auto-converge static 30ms throttling code and replace it
with a dynamic throttling algorithm.
Additionally, be more aggressive when deciding when to start throttling.
Previously we waited until four unproductive memory passes. Now we begin
throttling after only two unproductive memory passes. Four seemed quite
arbitrary and only waiting for two passes allows us to complete the migration
faster.
Jason J. Herne [Tue, 8 Sep 2015 17:12:34 +0000 (13:12 -0400)]
migration: Parameters for auto-converge cpu throttling
Add migration parameters to allow the user to adjust the parameters
that control cpu throttling when auto-converge is in effect. The added
parameters are as follows:
x-cpu-throttle-initial : Initial percantage of time guest cpus are throttled
when migration auto-converge is activated.
x-cpu-throttle-increment: throttle percantage increase each time
auto-converge detects that migration is not making progress.
Jason J. Herne [Tue, 8 Sep 2015 17:12:33 +0000 (13:12 -0400)]
cpu: Provide vcpu throttling interface
Provide a method to throttle guest cpu execution. CPUState is augmented with
timeout controls and throttle start/stop functions. To throttle the guest cpu
the caller simply has to call the throttle set function and provide a percentage
of throttle time.
* remotes/amit-migration/tags/for-juan-201509:
ram_find_and_save_block: Split out the finding
Move dirty page search state into separate structure
migration: Use g_new() & friends where that makes obvious sense
migration: qemu-file more size_t'ifying
migration: size_t'ify some of qemu-file
Init page sizes in qtest
Split out end of migration code from migration_thread
migration/ram.c: Use RAMBlock rather than MemoryRegion
vmstate: Remove redefinition of VMSTATE_UINT32_ARRAY
migration: Use g_new() & friends where that makes obvious sense
g_new(T, n) is neater than g_malloc(sizeof(T) * n). It's also safer,
for two reasons. One, it catches multiplication overflowing size_t.
Two, it returns T * rather than void *, which lets the compiler catch
more type errors.
This commit only touches allocations with size arguments of the form
sizeof(T). Same Coccinelle semantic patch as in commit b45c03f.
This is a start on using size_t more in qemu-file and friends;
it fixes up QEMUFilePutBufferFunc and QEMUFileGetBufferFunc
to take size_t lengths and return ssize_t return values (like read(2))
and fixes up all the different implementations of them.
Note that I've not yet followed this deeply into bdrv_ implementations.
Split out end of migration code from migration_thread
The code that gets run at the end of the migration process
is getting large, and I'm about to add more for postcopy.
Split it into a separate function.
migration/ram.c: Use RAMBlock rather than MemoryRegion
RAM migration mainly works on RAMBlocks but in a few places
uses data from MemoryRegions to access the same information that's
already held in RAMBlocks; clean it up just to avoid the
MemoryRegion use.
Peter Maydell [Fri, 25 Sep 2015 22:20:06 +0000 (23:20 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-cocoa-20150925-1' into staging
cocoa queue:
* fix stuck-key bug if keys were down when QEMU lost focus
* prompt the user whether they really meant to quit
* remove the 'open image file' dialog box we used to display
if the user started QEMU without arguments
* remotes/pmaydell/tags/pull-cocoa-20150925-1:
ui/cocoa.m: remove open dialog code
ui/cocoa.m: prevent stuck key situation
ui/cocoa.m: verify with user before quitting QEMU
John Arbuckle [Fri, 25 Sep 2015 22:14:00 +0000 (23:14 +0100)]
ui/cocoa.m: remove open dialog code
Removes the open dialog code that runs when no arguments are supplied with QEMU.
Not everyone needs a hard drive or cdrom to boot their target. A user might only
need to use their target's bios to do work. With that said, this patch removes
the unneeded open dialog code.
John Arbuckle [Fri, 25 Sep 2015 22:14:00 +0000 (23:14 +0100)]
ui/cocoa.m: prevent stuck key situation
When the user puts QEMU in the background while holding
down a key, QEMU will not receive the keyup event when
the user lets go of the key. When the user goes back to
QEMU, QEMU will think the key is still down causing
stuck key symptoms. This patch fixes this problem by
releasing all down keys when QEMU goes into the
background.
John Arbuckle [Fri, 25 Sep 2015 22:13:59 +0000 (23:13 +0100)]
ui/cocoa.m: verify with user before quitting QEMU
This patch prevents the user from accidentally quitting QEMU by pushing
Command-Q or by pushing the close button on the main window. When
the user does one of these two things, a dialog box appears verifying
with the user if he or she wants to quit QEMU.
Peter Maydell [Fri, 25 Sep 2015 20:52:30 +0000 (21:52 +0100)]
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
* First batch of MAINTAINERS updates
* IOAPIC fixes (to pass kvm-unit-tests with -machine kernel_irqchip=off)
* NBD API upgrades from Daniel
* strtosz fixes from Marc-André
* improved support for readonly=on on scsi-generic devices
* new "info ioapic" and "info lapic" monitor commands
* Peter Crosthwaite's ELF_MACHINE cleanups
* docs patches from Thomas and Daniel
# gpg: Signature made Fri 25 Sep 2015 11:20:52 BST using RSA key ID 78C7AE83
# gpg: Good signature from "Paolo Bonzini <[email protected]>"
# gpg: aka "Paolo Bonzini <[email protected]>"
* remotes/bonzini/tags/for-upstream: (52 commits)
doc: Refresh URLs in the qemu-tech documentation
docs: describe the QEMU build system structure / design
typedef: add typedef for QemuOpts
i386: interrupt poll processing
i386: partial revert of interrupt poll fix
ppc: Rename ELF_MACHINE to be PPC specific
i386: Rename ELF_MACHINE to be x86 specific
alpha: Remove ELF_MACHINE from cpu.h
mips: Remove ELF_MACHINE from cpu.h
sparc: Remove ELF_MACHINE from cpu.h
s390: Remove ELF_MACHINE from cpu.h
sh4: Remove ELF_MACHINE from cpu.h
xtensa: Remove ELF_MACHINE from cpu.h
tricore: Remove ELF_MACHINE from cpu.h
or32: Remove ELF_MACHINE from cpu.h
lm32: Remove ELF_MACHINE from cpu.h
unicore: Remove ELF_MACHINE from cpu.h
moxie: Remove ELF_MACHINE from cpu.h
cris: Remove ELF_MACHINE from cpu.h
m68k: Remove ELF_MACHINE from cpu.h
...
* remotes/cody/tags/block-pull-request:
sheepdog: refine discard support
sheepdog: use per AIOCB dirty indexes for non overlapping requests
Backup: don't do copy-on-read in before_write_notifier
block: Introduce a new API bdrv_co_no_copy_on_readv()
sheepdog: add reopen support
block/nfs: cache allocated filesize for read-only files
block/nfs: fix calculation of allocated file size
Peter Maydell [Fri, 25 Sep 2015 17:03:19 +0000 (18:03 +0100)]
Merge remote-tracking branch 'remotes/vivier-misc/tags/pull-muldiv64-20150925' into staging
Remove muldiv64() by using period instead of frequency
# gpg: Signature made Fri 25 Sep 2015 14:54:37 BST using RSA key ID 3F2FBE3C
# gpg: Good signature from "Laurent Vivier <[email protected]>"
# gpg: aka "Laurent Vivier <[email protected]>"
# gpg: aka "Laurent Vivier (Red Hat) <[email protected]>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: CD2F 75DD C8E3 A4DC 2E4F 5173 F30C 38BD 3F2F BE3C
Peter Maydell [Fri, 25 Sep 2015 15:40:05 +0000 (16:40 +0100)]
Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging
virtio,pc features, fixes
New features:
vhost-user multiqueue support
virtio-ccw virtio 1 support
Signed-off-by: Michael S. Tsirkin <[email protected]>
# gpg: Signature made Fri 25 Sep 2015 07:40:35 BST using RSA key ID D28D5469
# gpg: Good signature from "Michael S. Tsirkin <[email protected]>"
# gpg: aka "Michael S. Tsirkin <[email protected]>"
* remotes/mst/tags/for_upstream:
MAINTAINERS: add more devices to the PCI section
MAINTAINERS: add more devices to the PC section
vhost-user: add a new message to disable/enable a specific virt queue.
vhost-user: add multiple queue support
vhost: introduce vhost_backend_get_vq_index method
vhost-user: add VHOST_USER_GET_QUEUE_NUM message
vhost: rename VHOST_RESET_OWNER to VHOST_RESET_DEVICE
vhost-user: add protocol feature negotiation
vhost-user: use VHOST_USER_XXX macro for switch statement
virtio-ccw: enable virtio-1
virtio-ccw: feature bits > 31 handling
virtio-ccw: support ring size changes
virtio: ring sizes vs. reset
pc: Introduce pc-*-2.5 machine classes
q35: Move options common to all classes to pc_i440fx_machine_options()
q35: Move options common to all classes to pc_q35_machine_options()
virtio-net: unbreak self announcement and guest offloads after migration
virtio: right size for virtio_queue_get_avail_size
This patch refines discard support of the sheepdog driver. The
existing discard mechanism was implemented on SD_OP_DISCARD_OBJ, which
was introduced before fine grained reference counting on newer
sheepdog. It doesn't care about relations of snapshots and clones and
discards objects unconditionally.
With this patch, the driver just updates an inode object for updating
reference. Removing the object is done in sheep process side.
sheepdog: use per AIOCB dirty indexes for non overlapping requests
In the commit 96b14ff85acf, requests for overlapping areas are
serialized. However, it cannot handle a case of non overlapping
requests. In such a case, min_dirty_data_idx and max_dirty_data_idx
can be overwritten by the requests and invalid inode update can
happen e.g. a case like create(1, 2) and create(3, 4) are issued in
parallel.
This patch lets SheepdogAIOCB have dirty data indexes instead of
BDRVSheepdogState for avoiding the above situation.
This patch also does trivial renaming for better description:
overwrapping -> overlapping
Laurent Vivier [Tue, 25 Aug 2015 15:09:36 +0000 (17:09 +0200)]
arm: clarify the use of muldiv64()
muldiv64() is used to convert microseconds into CPU ticks.
But it is not clear and not commented. This patch uses macro
to clearly identify what is used: time, CPU frequency and ticks.
For an elapsed time and a given frequency, we compute how many ticks
we have.
Peter Lieven [Thu, 27 Aug 2015 10:30:41 +0000 (12:30 +0200)]
block/nfs: cache allocated filesize for read-only files
If the file is readonly its not expected to grow so
save the blocking call to nfs_fstat_async and use
the value saved at connection time. Also important
the monitor (and thus the main loop) will not hang
if block device info is queried and the NFS share
is unresponsive.
docs: describe the QEMU build system structure / design
Developers who are new to QEMU, or have a background familiarity
with GNU autotools, can have trouble getting their head around the
home-grown QEMU build system. This document attempts to explain
the structure / design of the configure script and the various
Makefile pieces that live across the source tree.
Pavel Dovgalyuk [Thu, 17 Sep 2015 16:24:11 +0000 (19:24 +0300)]
i386: interrupt poll processing
This patch updates x86_cpu_exec_interrupt function.
It can process two interrupt request at a time (poll and another one).
This makes its execution non-deterministic. Determinism is requred
for recorded icount execution.
Pavel Dovgalyuk [Thu, 17 Sep 2015 16:23:31 +0000 (19:23 +0300)]
i386: partial revert of interrupt poll fix
Processing CPU_INTERRUPT_POLL requests in cpu_has_work functions
break the determinism of cpu_exec. This patch is required to make
interrupts processing deterministic.
Rename ELF_MACHINE to be PPC specific. This is used as-is by the
various PPC bootloaders and is locally defined to ELF_MACHINE in linux
user in PPC specific ifdeffery.
This removes another architecture specific definition from the global
namespace (as desired by multi-arch).
The only generic code relying on this is linux-user. Linux user
already has a lot of #ifdef TARGET_ customisation so instead, define
ELF_ARCH as either EM_ARM or EM_AARCH64 appropriately.
The armv7m bootloader can just pass EM_ARM directly, as that
is architecture specific code. Note that arm_boot already has its own
logic selecting an arm specific elf machine so this makes V7M more
consistent with arm_boot.
This removes another architecture specific definition from the global
namespace.
linux-user: elfload: Provide default for elf_check_arch
For many arch's this macro is defined as the predicatable behaviour
of checking the argument for eqaulity against ELF_ARCH. Provide a
default define as such, so only archs with special handling (usually
allowing multiple EM values) need to provide a def.
Arches that do any of:
1: provide this def exactly the same way as the new default
(alpha, x86_64)
2: check against ELF_MACHINE while defining ELF_ARCH == ELF_MACHINE
(arm, aarch64)
3: check against EM_FOO directly while defining ELF_ARCH == EM_FOO
(unicore32, sparc32, ppc32, mips, openrisc, sh4, cris, m86k)
have their elf_check_arch removed as the default will provide the
correct behaviour.
linux_user: elfload: Default ELF_MACHINE to ELF_ARCH
In most (but not all) cases, ELF_MACHINE and ELF_ARCH are safely the
same. Default ELF_MACHINE to ELF_ARCH. This makes defining ELF_MACHINE
optional for target-*/cpu.h when they are known to match.
Pavel Butsykin [Tue, 22 Sep 2015 13:18:16 +0000 (16:18 +0300)]
apic_internal.h: fix formatting and drop unused consts
Fix formatting of local apic definitions and drop unused constant
APIC_INPUT_POLARITY, APIC_SEND_PENDING. Magic numbers in shifts are
replaced with constants defined just above.
Pavel Butsykin [Tue, 22 Sep 2015 13:18:13 +0000 (16:18 +0300)]
apic_internal.h: make some apic_get_* functions externally visible
Move apic_get_bit(), apic_set_bit() to apic_internal.h, make the apic_get_ppr
symbol external. It's necessary to work with isr, tmr, irr and ppr outside
hw/intc/apic.c
Paolo Bonzini [Thu, 30 Jul 2015 08:19:24 +0000 (10:19 +0200)]
ioapic: coalesce level interrupts
If a level-triggered interrupt goes down and back up before the
corresponding EOI, it should be coalesced. This fixes one testcase
in kvm-unit-tests' ioapic.flat.
Paolo Bonzini [Tue, 22 Sep 2015 09:54:32 +0000 (11:54 +0200)]
get_maintainer.pl: \C is deprecated
"Match a single C-language char (octet) even if that is part of a larger
UTF-8 character. Thus it breaks up characters into their UTF-8 bytes,
so you may end up with malformed pieces of UTF-8."
Makefile: fix build when VPATH is outside GIT tree
Steve Ellcey / Leon Alrae reported that QEMU fails to build when
the VPATH directory is outside of the GIT tree, and the system
emulators & tools build is disabled. eg
cd ..
mkdir build
cd build
../qemu/configure --disable-system --disable-tools
make
(...)
make[1]: *** No rule to make target `../qom/object.o', needed by `qemu-aarch64'. Stop.
make: *** [subdir-aarch64-linux-user] Error 2
The problem is due to the fact that some sub directory deps
were listed against SOFTMMU_SUBDIR_RULES instead of SUBDIR_RULES,
so were only processed for system emulators, not user emalutors.
Paolo Bonzini [Wed, 16 Sep 2015 15:26:16 +0000 (17:26 +0200)]
scsi-generic: let guests recognize readonly=on on passthrough devices
Passed-through SCSI devices can be opened with the readonly=on option.
When this happens, Linux filters away write commands so that the guest
cannot overwrite the contents of the device.
However, the guest does not know that the device is read-only, and
accepts writes. The writes only fail later when the page cache is
flushed.
This patch modifies scsi-generic to modify the MODE SENSE data and
set the read-only bit in the device-specific parameters, so that
the guest OS treats the disk as write protected.
projects / qemu.git / log