Alberto Garcia [Thu, 15 Sep 2016 14:53:02 +0000 (17:53 +0300)]
block: Add "read-only" to the options QDict
This adds the "read-only" option to the QDict. One important effect of
this change is that when a child inherits options from its parent, the
existing "read-only" mode can be preserved if it was explicitly set
previously.
This addresses scenarios like this:
[E] <- [D] <- [C] <- [B] <- [A]
In this case, if we reopen [D] with read-only=off, and later reopen
[B], then [D] will not inherit read-only=on from its parent during the
bdrv_reopen_queue_child() stage.
The BDRV_O_RDWR flag is not removed yet, but its keep in sync with the
value of the "read-only" option.
Alberto Garcia [Thu, 15 Sep 2016 14:53:01 +0000 (17:53 +0300)]
block: Update bs->open_flags earlier in bdrv_open_common()
We're only doing this immediately before opening the image, but
bs->open_flags is used earlier in the function. At the moment this is
not causing problems because none of the checked flags are modified by
update_flags_from_options(), but this will change when we introduce
the "read-only" option.
This patch calls update_flags_from_options() at the beginning of the
function, immediately after creating the QemuOpts.
Alberto Garcia [Thu, 15 Sep 2016 14:53:00 +0000 (17:53 +0300)]
block: Set BDRV_O_ALLOW_RDWR and snapshot_options before storing the flags
If an image is opened with snapshot=on, its flags are modified by
bdrv_backing_options() and then bs->open_flags is updated accordingly.
This last step is unnecessary if we calculate the new flags before
setting bs->open_flags.
Soon we'll introduce the "read-only" option, and then we'll need to
be able to modify its value in the QDict when snapshot=on. This is
more cumbersome if bs->options is already set. This patch simplifies
that. Other than that, there are no semantic changes. Although it
might seem that bs->options can have a different value now because
it is stored after calling bdrv_backing_options(), this call doesn't
actually modify them in this scenario.
The code that sets BDRV_O_ALLOW_RDWR is also moved for the same
reason.
The copy_sectors() code was originally using the 'sector'
parameter for encryption, which was passed in by the caller
from the QCowL2Meta.offset field (aka the guest logical
offset).
After the change, the code is using 'cluster_offset' which
was passed in from QCow2L2Meta.alloc_offset field (aka the
host physical offset).
This would cause the data to be encrypted using an incorrect
initialization vector which will in turn cause later reads
to return garbage.
Although current qcow2 built-in encryption is blocked from
usage in the emulator, one could still hit this if writing
to the file via qemu-{img,io,nbd} commands.
* remotes/bonzini/tags/for-upstream:
kvm: fix events.flags (KVM_VCPUEVENT_VALID_SMM) overwritten by 0
scripts: Add a script to check for bug URLs in the git log
msmouse: Fix segfault caused by free the chr before chardev cleanup.
iscsi: Fix divide-by-zero regression on raw SG devices
kvm: apic: set APIC base as part of kvm_apic_put
target-i386: introduce kvm_put_one_msr
Such output may be further improved by providing strace-related functions
that handle only particular syscalls, but this is beyond the scope of
this patch.
linux-user: Add missing TARGET_EDQUOT error code for Mips
EDQUOT is defined for Mips platform in Linux kernel in such a way
that it has different value than on most other platforms. However,
correspondent TARGET_EDQUOT for Mips is missing in Qemu code. Moreover,
TARGET_EDQUOT is missing from the table for conversion of error codes
from host to target. This patch fixes these problems.
Without this patch, syscalls add_key(), keyctl(), link(), mkdir(), mknod(),
open(), rename(), request_key(), setxattr(), symlink(), and write() will not
be able to return the right error code in some scenarios on Mips platform.
(Some of these syscalls are not yet supported in Qemu, but once they are
supported, they will need correct EDQUOT handling.)
linux-user: Fix certain argument alignment cases for Mips64
The function that is changed in this patch is supposed to indicate that
there was certain argument rearrangement related to 64-bit arguments on
32-bit platforms. The background on such rearrangements can be found,
for example, in the man page for syscall(2).
However, for 64-bit Mips architectures there is no such rearrangement,
and this patch reflects it.
linux-user: Fix structure target_flock definition for Mips
Structure flock is defined for Mips in a way different from any
other platform. For reference, see Linux kernel source code files:
arch/mips/include/uapi/asm/fcntl.h, line 63 (for Mips)
include/uapi/asm-generic/fcntl.h, line 195 (for all other platforms)
This patch fix this problem, by amending structure target_flock,
for Mips only.
Besides, this patch fixes LTP tests fcntl11, fcntl17, fcntl19, fcntl20,
and fcntl21, which are currently failing, if executed in Qemu user mode
for Mips platforms.
André Draszik [Mon, 25 Jul 2016 23:42:45 +0000 (00:42 +0100)]
target-mips: add 24KEc CPU definition
Define a new CPU definition supporting 24KEc cores, similar to
the existing 24Kc, but with added support for DSP instructions
and MIPS16e (and without FPU).
sscanf is relatively loose (tolerate) on some invalid formats that we
should fail instead of generating a wrong uuid structure, like with
whitespaces and short strings.
Add and use a helper function to first check the format.
This removes our dependency to libuuid, so that the driver can always be
built.
Similar to how we handled data plane configure options, --enable-vhdx
and --disable-vhdx are also changed to a nop with a message saying it's
obsolete.
A number of different places across the code base use CONFIG_UUID. Some
of them are soft dependency, some are not built if libuuid is not
available, some come with dummy fallback, some throws runtime error.
It is hard to maintain, and hard to reason for users.
Since UUID is a simple standard with only a small number of operations,
it is cleaner to have a central support in libqemuutil. This patch adds
qemu_uuid_* functions that all uuid users in the code base can
rely on. Except for qemu_uuid_generate which is new code, all other
functions are just copy from existing fallbacks from other files.
Note that qemu_uuid_parse is moved without updating the function
signature to use QemuUUID, to keep this patch simple.
docker: Handle exceptions when looking for docker command
When trying to run docker tests on a host without the docker
command, we get the following Python backtrace:
$ make docker-test-quick@centos6 V=1
.../qemu/tests/docker/docker.py build qemu:centos6 .../qemu/tests/docker/dockerfiles/centos6.docker
Traceback (most recent call last):
File ".../qemu/tests/docker/docker.py", line 339, in <module>
sys.exit(main())
File ".../qemu/tests/docker/docker.py", line 336, in main
return args.cmdobj.run(args, argv)
File ".../qemu/tests/docker/docker.py", line 231, in run
dkr = Docker()
File ".../qemu/tests/docker/docker.py", line 98, in __init__
self._command = _guess_docker_command()
File ".../qemu/tests/docker/docker.py", line 41, in _guess_docker_command
stdout=DEVNULL, stderr=DEVNULL) == 0:
File "/usr/lib64/python2.7/subprocess.py", line 523, in call
return Popen(*popenargs, **kwargs).wait()
File "/usr/lib64/python2.7/subprocess.py", line 711, in __init__
errread, errwrite)
File "/usr/lib64/python2.7/subprocess.py", line 1343, in _execute_child
raise child_exception
OSError: [Errno 2] No such file or directory
.../qemu/tests/docker/Makefile.include:47: recipe for target 'docker-image-centos6' failed
make: *** [docker-image-centos6] Error 1
Change _guess_docker_command() to handle OSError exceptions
raised by subprocess.call(), so we will keep looking for other
commands and print a better error message.
New output will be:
$ make docker-test-quick@centos6 V=1
.../qemu/tests/docker/docker.py build qemu:centos6 .../qemu/tests/docker/dockerfiles/centos6.docker
Traceback (most recent call last):
File ".../qemu/tests/docker/docker.py", line 343, in <module>
sys.exit(main())
File ".../qemu/tests/docker/docker.py", line 340, in main
return args.cmdobj.run(args, argv)
File ".../qemu/tests/docker/docker.py", line 235, in run
dkr = Docker()
File ".../qemu/tests/docker/docker.py", line 102, in __init__
self._command = _guess_docker_command()
File ".../qemu/tests/docker/docker.py", line 49, in _guess_docker_command
commands_txt)
Exception: Cannot find working docker command. Tried:
docker
sudo -n docker
.../qemu/tests/docker/Makefile.include:47: recipe for target 'docker-image-centos6' failed
make: *** [docker-image-centos6] Error 1
This adds a numa id property to a PHB to allow linking passed PCI device
to CPU/memory. It is up to the management stack to do CPU/memory pinning
to the node with the actual PCI device.
Signed-off-by: Alexey Kardashevskiy <[email protected]>
[dwg: Renamed property from "node" to "numa_node" to match the similar
one in the pxb device] Signed-off-by: David Gibson <[email protected]>
David Gibson [Wed, 21 Sep 2016 05:29:26 +0000 (15:29 +1000)]
monitor: fix crash for platforms without a CPU 0
Now that we allow CPU hot unplug on a few platforms, we can end up in a
situation where we don't have a CPU with index 0. Or at least we could,
if we didn't have code to explicitly prohibit unplug of CPU 0.
Longer term we want to allow CPU 0 unplug, this patch is an early step in
allowing this, by removing an assumption in the monitor code that CPU 0
always exists.
Michael Walle [Wed, 21 Sep 2016 09:57:05 +0000 (11:57 +0200)]
linux-user: ppc64: fix ARCH_206 bit in AT_HWCAP
Only the POWER[789] CPUs should have the ARCH_206 bit set. This is what the
linux kernel does. I guess this was also the intention of commit 0e019746.
We have to make sure all *206 bits are set.
Before this patch, the flags check in the GET_FEATURES2 macro returned true
if _any_ bit was set. This worked well as long as there was only one bit
set in the 'flag' parameter. But as explained before, we have to make sure
all bits in the 'flag' parameter are set.
Thomas Huth [Wed, 21 Sep 2016 09:42:15 +0000 (11:42 +0200)]
ppc/kvm: Mark 64kB page size support as disabled if not available
QEMU currently refuses to start with KVM-PR and only prints out
qemu: fatal: Unknown MMU model 851972
when being started there. This is because commit 4322e8ced5aaac719
("ppc: Fix 64K pages support in full emulation") introduced a new
POWERPC_MMU_64K bit to indicate support for this page size, but
it never gets cleared on KVM-PR if the host kernel does not support
this. Thus we've got to turn off this bit in the mmu_model for KVM-PR.
Fix inconsistent irq status, because of this in the trace logs, for e.g.
LSI status was 0x7, i.e. XICS_STATUS_ASSERTED, XICS_STATUS_SENT and
XICS_STATUS_REJECTED all set, which did not make sense. So the REJECTED
would have been set in earlier interrupt cycle, and then asserted and
sent in this current one.
Nathan Whitehorn [Tue, 30 Aug 2016 01:02:47 +0000 (01:02 +0000)]
Enable H_CLEAR_MOD and H_CLEAR_REF hypercalls on KVM/PPC64.
These are mandatory per PAPR and available on Linux 4.3 and newer kernels. The calls in question are required to run FreeBSD guests with reasonable performance, so enable them if possible.
Signed-off-by: Nathan Whitehorn <[email protected]>
[dwg: Added a stub to fix compile without KVM (e.g. on x86 host)] Signed-off-by: David Gibson <[email protected]>
target-ppc: tlbie/tlbivax should have global effect
tlbie (BookS) and tlbivax (BookE) plus the H_CALLs(pseries) should have
a global effect.
Introduces TLB_NEED_GLOBAL_FLUSH flag. During lazy tlb flush, after
taking care of pending local flushes, check broadcast flush(at context
synchronizing event ptesync/tlbsync, etc) is needed. Depending on the
bitmask state of the tlb_need_flush, tlb is flushed from other cpus if
needed and the flags are cleared.
We flush the qemu TLB lazily. check_tlb_flush is called whenever we hit
a context synchronizing event or instruction that requires a pending
flush to be performed.
However, we fail to handle broadcast TLB flush operations. In order to
fix that efficiently, we want to differentiate whether check_tlb_flush()
needs to only apply pending local flushes (isync instructions,
interrupts, ...) or also global pending flush operations. The latter is
only needed when executing instructions that are defined architecturally
as synchronizing global TLB flush operations. This in our case is
ptesync on BookS and tlbsync on BookE along with the paravirtualized
hypervisor calls.
Signed-off-by: Nikunj A Dadhania <[email protected]>
[dwg: Changed gen_check_tlb_flush() to also take a bool, and fixed
some spelling errors in commit message] Signed-off-by: David Gibson <[email protected]>
Each spapr cpu core type defines an instance_init routine which just
populates the CPU class name. This can be done in the class_init
commonly for all core types which simplifies the registration.
This is inspired by how PowerNV core types are registered.
Certain types of spapr cpu cores ('host' and generic type based on host
CPU) are initialized in target-ppc/kvm.c. To convert these type
registrations to use class_init, we need to expose
spapr_cpu_core_class_init() outside of spapr_cpu_core.c.
Commit d11b268e1765 added a generic sPAPR CPU core family
type to support cases like POWER8 CPU type on POWER8E host CPU.
Switching to class_init would fix such scenarios to use the right
CPU thread type instead of defaulting to host-powerpc64-cpu.
In an unrelated cleanup, fix a typo in .get_hotplug_handler routine.
Ravi Bangoria [Fri, 16 Sep 2016 10:51:47 +0000 (16:21 +0530)]
target-ppc: implement darn instruction
darn: Deliver A Random Number
Currently return invalid random number for all the case. This needs
proper algorithm to provide cryptographically suitable random data.
Reading from /dev/random can block and that is not an expected behaviour
while the cpu instruction is getting executed. Moreover, /dev/random
would only work for linux-user
Being a 16byte operation, qemu_ld/st still does not support this. Move
this out so other store operation can use qemu_ld/st in the following
patch. Also, convert it to two MO_Q operations for stqcx.
Power ISA 2.x has deleted the rfi instruction and rfid shoud be used
instead on cpus following this instruction set or later.
This will raise an invalid exception when rfi is used on such
processors: Book3S 64-bit processors.
Signed-off-by: Benjamin Herrenschmidt <[email protected]> Reviewed-by: David Gibson <[email protected]>
[clg: the required fix in openbios, commit b747b6acc272 ('ppc: use
rfid when running under a CPU from the 970 family.'), is now
merged in qemu under commit 5cebd885d0d2 ('Update OpenBIOS
images to b747b6a built from submodule.') ] Signed-off-by: Cédric Le Goater <[email protected]> Reviewed-by: Thomas Huth <[email protected]> Signed-off-by: David Gibson <[email protected]>
Thomas Huth [Mon, 5 Sep 2016 20:03:36 +0000 (22:03 +0200)]
MAINTAINERS: Add some missing ppc-related files
There are some powerpc related files in the QEMU source tree
which are currently not covered by the MAINTAINERS file and
thus not properly classified by the get_maintainer.pl script.
So let's add them to the proper sections.
Thomas Huth [Wed, 21 Sep 2016 19:42:22 +0000 (21:42 +0200)]
scripts: Add a script to check for bug URLs in the git log
Basic idea of this script is to check the git log for URLs
to the QEMU bugtracker at launchpad.net and to figure out
whether the related bug has been marked there as "Fix released"
(i.e. closed) already. So this script can e.g. be used after
each public release of QEMU to check whether there are any
bug tickets that could be moved from "Fix committed" (or another
state if the author of the patch forgot to update the bug ticket)
to "Fix released".
Lin Ma [Thu, 15 Sep 2016 14:31:58 +0000 (22:31 +0800)]
msmouse: Fix segfault caused by free the chr before chardev cleanup.
Segfault happens when leaving qemu with msmouse backend:
#0 0x00007fa8526ac975 in raise () at /lib64/libc.so.6
#1 0x00007fa8526add8a in abort () at /lib64/libc.so.6
#2 0x0000558be78846ab in error_exit (err=16, msg=0x558be799da10 ...
#3 0x0000558be7884717 in qemu_mutex_destroy (mutex=0x558be93be750) at ...
#4 0x0000558be7549951 in qemu_chr_free_common (chr=0x558be93be750) at ...
#5 0x0000558be754999c in qemu_chr_free (chr=0x558be93be750) at ...
#6 0x0000558be7549a20 in qemu_chr_delete (chr=0x558be93be750) at ...
#7 0x0000558be754a8ef in qemu_chr_cleanup () at qemu-char.c:4643
#8 0x0000558be755843e in main (argc=5, argv=0x7ffe925d7118, ...
The chr was freed by msmouse close callback before chardev cleanup,
Then qemu_mutex_destroy triggered raise().
Because freeing chr is handled by qemu_chr_free_common, Remove the free from
msmouse_chr_close to avoid double free.
Eric Blake [Wed, 7 Sep 2016 21:27:20 +0000 (16:27 -0500)]
iscsi: Fix divide-by-zero regression on raw SG devices
When qemu uses iscsi devices in sg mode, iscsilun->block_size
is left at 0. Prior to commits cf081fca and similar, when
block limits were tracked in sectors, this did not matter:
various block limits were just left at 0. But when we started
scaling by block size, this caused SIGFPE.
Then, in a later patch, commit a5b8dd2c added an assertion to
bdrv_open_common() that request_alignment is always non-zero;
which was not true for SG mode. Rather than relax that assertion,
we can just provide a sane value (we don't know of any SG device
with a block size smaller than qemu's default sizing of 512 bytes).
One possible solution for SG mode is to just blindly skip ALL
of iscsi_refresh_limits(), since we already short circuit so
many other things in sg mode. But this patch takes a slightly
more conservative approach, and merely guarantees that scaling
will succeed, while still using multiples of the original size
where possible. Resulting limits may still be zero in SG mode
(that is, we mostly only fix block_size used as a denominator
or which affect assertions, not all uses).
Peter Maydell [Thu, 22 Sep 2016 17:23:14 +0000 (18:23 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20160922' into staging
target-arm queue:
* add Cortex-A7 CPU
* new ast2500 SoC model and evaluation board
* palmetto-bmc: remove stray double assignment
* aspeed: clean up RAM size handling
* ptimer: framework for defining policy bits to change
behaviour choices for different timer devices
* ptimer: add some test cases
* cadence_gem: add queue support
* loader: support loading images to specified address spaces
* loader: support auto-detect of ELF architecture from file
* dma: xlnx-zynq-devcfg: Fix up XLNX_ZYNQ_DEVCFG_R_MAX
* vmstateify ssd0323
* vmstateify ssi-sd
* disas/arm.c: remove unused macros
* imx: use 'const char', not 'char const'
* remotes/pmaydell/tags/pull-target-arm-20160922: (36 commits)
imx: Use 'const char', not 'char const'
disas/arm.c: Remove unused macro definitions
vmstateify ssi-sd
vmstateify ssd0323 display
dma: xlnx-zynq-devcfg: Fix up XLNX_ZYNQ_DEVCFG_R_MAX
loader: Add AddressSpace loading support to targphys
loader: Add AddressSpace loading support to uImages
loader: Add AddressSpace loading support to ELFs
loader: Allow a custom AddressSpace when loading ROMs
loader: Use the specified MemoryRegion
loader: Allow ELF loader to auto-detect the ELF arch
xlnx-zynqmp: Set the number of priority queues
cadence_gem: Correct indentation
cadence_gem: Add queue support
cadence_gem: Add support for screening
cadence_gem: Add the num-priority-queues property
cadence_gem: QOMify Cadence GEM
tests: Add ptimer tests
hw/ptimer: Suppress error messages under qtest
hw/ptimer: Introduce timer policy feature
...
Peter Maydell [Thu, 22 Sep 2016 17:13:09 +0000 (18:13 +0100)]
imx: Use 'const char', not 'char const'
'char const' means the same thing as 'const char', but we
use the former in only a handful of places and we use the
latter over six thousand times. Switch the imx reg_name()
functions to bring them in line with everything else.
Changed a few types to fixed sized types in the ssi_sd_state
Now saving/loading a byte for the cmdarg/response bytes that were
previously saved as uint32
Bumped version number to deal with those changes.
dma: xlnx-zynq-devcfg: Fix up XLNX_ZYNQ_DEVCFG_R_MAX
Whilst according to the Zynq TRM this device covers a register region of
0x000 - 0x120. The register region is also shared with XADCIF prefix
registers at 0x100 and above. Due to how the devcfg and the xadc devices
are implemented in QEMU these are separate models with individual mmio
regions. As such the region registered by the devcfg overlaps with the
xadc when initialized in a machine model (e.g. xilinx-zynq-a9).
This patch fixes up the incorrect region size, where
XLNX_ZYNQ_DEVCFG_R_MAX is missing its '/ 4' causing it to be 0x460 in
size. As well as setting the region size to the 0x0 - 0x100 region so
that an xadc device instance can be registered in the correct region to
pair with the devcfg device instance.
Alistair Francis [Thu, 22 Sep 2016 17:13:08 +0000 (18:13 +0100)]
loader: Add AddressSpace loading support to targphys
Add a new function load_image_targphys_as() that allows the caller
to specify an AddressSpace to use when loading a targphys. The
original load_image_targphys() function doesn't have any change in
functionality.
Alistair Francis [Thu, 22 Sep 2016 17:13:08 +0000 (18:13 +0100)]
loader: Add AddressSpace loading support to uImages
Add a new function load_uimage_as() that allows the caller to
specify an AddressSpace to use when loading the uImage. The
original load_uimage() function doesn't have any change in
functionality.
Alistair Francis [Thu, 22 Sep 2016 17:13:08 +0000 (18:13 +0100)]
loader: Add AddressSpace loading support to ELFs
Add a new function load_elf_as() that allows the caller to specify an
AddressSpace to use when loading the ELF. The original load_elf()
function doesn't have any change in functionality.
projects / qemu.git / log