Define OPC_CALL_Jz, generated by tcg_out_calli; use the later
throughout. Unify the calls within qemu_st; adjust the stack
with a single pop if applicable.
Add more OPC values, and tgen_arithr. Use the later throughout.
Note that normal reg/reg arithmetic now uses the Gv,Ev opcode form
instead of the Ev,Gv opcode form used previously. Both forms
disassemble properly, and so there's no visible change when diffing
log files before and after the change. This change makes the operand
ordering within the output routines more natural, and avoids the need
to define an OPC_ARITH_EvGv since a read-modify-write with memory is
not needed within TCG.
Rabin Vincent [Sun, 2 May 2010 09:50:51 +0000 (15:20 +0530)]
arm_timer: reload timer when enabled
Reload the timer when TimerControl is written, if the timer is to be
enabled. Otherwise, if an earlier write to TimerLoad was done while
periodic mode was not set, s->delta may incorrectly still have the value
of the maximum limit instead of the value written to TimerLoad.
This problem is evident on versatileap on current linux-next, which
enables TIMER_CTRL_32BIT before writing to TimerLoad and then enabling
periodic mode and starting the timer. This causes the first periodic
tick to be scheduled to occur after 0xffffffff periods, leading to a
perceived hang while the kernel waits for the first timer tick.
target-sparc: Inline some generation of carry for ADDX/SUBX.
Computing carry is trivial for some inputs. By avoiding an
external function call, we generate near-optimal code for
the common cases of add+addx (double-word arithmetic) and
cmp+addx (a setcc pattern).
If the address register overlaps one of the output registers
simply issue the clobbering load last, rather than emitting
an extra move of the address register.
Use int32 types instead of target_ulong when computing ICC. This
simplifies the generated code for 32-bit host and 64-bit guest.
Use the same simplified expressions for ICC as were already used
for XCC in carry flag generation.
Simplify the ADD carry generation to not consider a possible carry-in.
Use the more complex carry computation for ADDX only. Use the same
carry algorithm for the XCC result of ADDX. Similarly for SUB/SUBX.
Use the ADD carry generation functions for TADD/TADDTV. Similarly
for SUB and TSUB/TSUBTV.
Stefan Weil [Sun, 11 Apr 2010 16:44:18 +0000 (18:44 +0200)]
Fix cross compilation
This patch enhances the algorithm which finds the correct settings for SDL.
For cross compilations (when cross_prefix is set), it looks for sdl-config
with cross prefix. Here is the complete search order:
$(cross_prefix}pkg-config (old, only used for cross compilation)
${cross_prefix}sdl_config (new, only used for cross compilation)
pkg-config (old, needs PATH)
sdl-config (old, needs PATH)
Cross SDL packages (or the user) now can simply set a link (for example
/usr/bin/i586-mingw32msvc-sdl-config -> /usr/i586-mingw32msvc/bin/sdl-config)
which allows cross compilations without PATH modifications.
Without the patch, configure and make (which calls configure) typically
need a non-standard PATH. Failing to set this special PATH results in
broken builds.
Alexander Graf [Fri, 14 May 2010 14:14:31 +0000 (16:14 +0200)]
target-s390: enable SIGP Initial Reset
For SMP to work with KVM, we need to properly emulate the SIGP Initial Reset
Command. Recent (2.6.32) kernels issue that before the SIGP Reset command that
actually wakes up the vcpu.
Alexander Graf [Tue, 20 Apr 2010 17:37:13 +0000 (19:37 +0200)]
target-s390: add firmware code
This patch adds a firmware blob to the S390 target. The blob is a simple
implementation of a virtio client that tries to read the second stage
bootloader from sectors described as of offset 0x20 in the MBR.
In combination with an updated zipl this allows for booting from virtio
block devices. This firmware is built from the same sources as the second
stage bootloader. You can find a virtio capable s390-tools in this repo:
Alexander Graf [Sun, 18 Apr 2010 21:10:17 +0000 (23:10 +0200)]
PPC/KVM: make iothread work
When running with --enable-io-thread the timer we have doesn't help,
because it doesn't wake up the CPU thread. So instead we need to
actually kick it.
While at it I refined the logic a bit to not dumbly trigger a timer
every 500ms, but rather do it more often after an interrupt got injected.
If there's no level based interrupt to be expected, we don't need the
timer anyways.
This makes qemu-system-ppc with --enable-io-thread work when using KVM.
If the user wants to create a chardev of type socket but forgets to give a
host= option, qemu_opt_get returns NULL. This NULL pointer is then fed into
strlen a few lines below without a check which results in a segfault.
This fixes it.
linux-user: rlimit conversion between host and target.
rlim_t conversion between host and target added.
Otherwise there are some incorrect case like
- RLIM_INFINITY on 32bit target -> 64bit host.
- RLIM_INFINITY on 64bit host -> mips and sparc target ?
- Big value(for 32bit target) on 64bit host -> 32bit target.
One is added into getrlimit, setrlimit, and ugetrlimit. It converts both
RLIM_INFINITY and value bigger than target can hold(>31bit) to RLIM_INFINITY.
Another one is added to guest_stack_size calculation introduced by 703e0e89. The rule is mostly same except the result on the case is keeping
the value of guest_stack_size.
Slightly tested for SH4, and x86_64 -linux-user on x86_64-pc-linux host.
Kevin Wolf [Wed, 12 May 2010 12:03:02 +0000 (14:03 +0200)]
block: Remove special case for vvfat
The special case doesn't really us buy anything. Without it vvfat works more
consistently as a protocol. We get raw on top of vvfat now, which works just
as well as using vvfat directly.
Stefan Weil [Wed, 12 May 2010 18:25:45 +0000 (20:25 +0200)]
block/vdi: Fix image opening and creation for odd disk sizes
The fix is based on a patch from Kevin Wolf. Here his comment:
"The number of blocks needs to be rounded up to cover all of the virtual hard
disk. Without this fix, we can't even open our own images if their size is not
a multiple of the block size."
While Kevin's patch addressed vdi_create, my modification also fixes
vdi_open which now accepts images with odd disk sizes.
v3:
Don't allow reading of disk images with too large disk sizes.
Neither VBoxManage nor old versions of qemu-img read such images.
This change requires rounding of odd disk sizes before we do the checks.
Use bdrv_pwrite to access the backing device instead of pread, and
convert the driver to implementing the bdrv_open method which gives
it an already opened BlockDriverState for the underlying device.
Dmg actually does an lseek to a negative offset in the open routine,
which we replace with offset arithmetics after doing a bdrv_getlength.
Use pread instead of lseek + read in preparation of using the qemu
block API. Note that dmg actually uses the implicit file offset
a lot in dmg_open, and we had to replace it with an offset variable.
When dmg_read_chunk encounters an uncompressed chunk it currently
calls read without any previous adjustment of the file postion.
This seems very wrong, and the "reference" implementation in
dmg2img does a search to the same offset as done in the various
compression cases, so do the same here.
Use bdrv_pwrite to access the backing device instead of pread, and
convert the driver to implementing the bdrv_open method which gives
it an already opened BlockDriverState for the underlying device.
Kevin Wolf [Thu, 6 May 2010 14:34:56 +0000 (16:34 +0200)]
block: Fix bdrv_commit
When reopening the image, don't guess the driver, but use the same driver as
was used before. This is important if the format=... option was used for that
image.
Use bdrv_pwrite to access the backing device instead of pread, and
convert the driver to implementing the bdrv_open method which gives
it an already opened BlockDriverState for the underlying device.
Kevin Wolf [Tue, 4 May 2010 14:35:24 +0000 (16:35 +0200)]
ide: Fix ide_dma_cancel
When cancelling a request, bdrv_aio_cancel may decide that it waits for
completion of a request rather than for cancellation. IDE therefore can't
abandon its DMA status before calling bdrv_aio_cancel; otherwise the callback
of a completed request would use invalid data.
Use bdrv_pwrite to access the backing device instead of pread, and
convert the driver to implementing the bdrv_open method which gives
it an already opened BlockDriverState for the underlying device.
Ryota Ozaki [Sun, 2 May 2010 21:50:25 +0000 (06:50 +0900)]
qemu-nbd: Improve error reporting
- use err(3) instead of errx(3) if errno is available
to report why failed
- let fail prior to daemon(3) if opening a nbd file
is likely to fail after daemonizing to avoid silent
failure exit
- add missing 'ret = 1' when unix_socket_outgoing failed
Isaku Yamahata [Fri, 14 May 2010 07:29:09 +0000 (16:29 +0900)]
pc: make pc_init1() not refer ferr_irq directly.
By introducing a registering function, make pc_init1() not refer to
ferr_irq directly in order to make ferr_irq piix independent.
Later pc_init1() will be split out into another file keeping ferr_irq
static.
Isaku Yamahata [Fri, 14 May 2010 07:29:08 +0000 (16:29 +0900)]
pc: introduce a function to allocate cpu irq.
Introduce a function, pc_allocate_cpu_irq(), to allocate cpu irq
in order to make pic_irq_request() piix independent.
Later piix code will be split out to another file keeping pic_irq_request()
static.
Isaku Yamahata [Fri, 14 May 2010 07:29:03 +0000 (16:29 +0900)]
pc: initialize ioapic before use.
The changeset of 2c8d9340203c7f19265fd4cb2341f568217a3af6
prevents isa_irq_handler() from NULL refering of IsaIrqState::ioapic.
However it would be better to initialize the member before reference.
Jan Kiszka [Sat, 15 May 2010 11:32:41 +0000 (13:32 +0200)]
serial: Register vmstate via qdev
At least for isa-serial, we can already let qdev do the vmstate
registration for us. It just takes wrapping vmstate for the
encapsulating ISASerialState and defining the proper instance ID
aliases.
Jan Kiszka [Sat, 15 May 2010 11:32:40 +0000 (13:32 +0200)]
vmstate: Add support for alias ID
Some legacy users (mostly PC devices) of vmstate_register manage
instance IDs on their own, and that unfortunately in a way that is
incompatible with automatically generated ones. This so far prevents
switching those users to vmstates that are registered by qdev.
To establish a migration path, this patch introduces the concept of
alias IDs. They can be passed to an extended vmstate registration
service, and qdev provides a set service to be used during device init.
find_se will consider the alias in addition to the default ID. We can
then start generating the default ID automatically and writing it on
vmsave, thus converting that format without breaking support for upward
migration.
The user is required specify the highest vmstate version for which the
alias is required. Once this version falls behind the minimum required
for a specific vmstate, an assertion triggers to motivate cleaning up
the obsolete alias.
projects / qemu.git / log