Bruce Rogers [Tue, 27 Nov 2012 20:11:25 +0000 (13:11 -0700)]
qapi: fix qapi_dealloc_type_size parameter type
The second parameter to qapi_dealloc_type_size should be a uint64_t *,
not a size_t *. This was causing our 32 bit x86 build to fail, since
warnings are treated as errors.
Anthony Liguori [Wed, 28 Nov 2012 19:19:24 +0000 (13:19 -0600)]
Merge remote-tracking branch 'bonzini/scsi-next' into staging
* bonzini/scsi-next:
virtio-scsi: Fix subtle (guest) endian bug
virtio-scsi: Fix some endian bugs with virtio-scsi
iscsi: do not assume device is zero initialized
iscsi: fix deadlock during login
iscsi: fix segfault in url parsing
Paolo Bonzini [Wed, 28 Nov 2012 08:40:23 +0000 (09:40 +0100)]
build: compile translate.o with -fno-gcse option on GCC 4.6.x and 4.7.[012]
These versions of GCC require insane (>2GB) amounts of memory to compile
translate.o. As a countermeasure, disable the culprit optimization pass.
This should fix the buildbot failure for default_x86_64_fedora16. Anyway
this is a good thing to do because people will try to compile 1.3 with
less than 2GB of memory and complain.
David Gibson [Mon, 26 Nov 2012 01:33:52 +0000 (12:33 +1100)]
virtio-scsi: Fix subtle (guest) endian bug
The virtio-scsi config space is, by specification, in guest endian (which
is ill-defined, but there you go). In virtio_scsi_get_config() we set up
all the fields in there, using stl_raw(). Which is a problem for the
max_channel and max_target fields, which are 16-bit, not 32-bit. For
little-endian targets we get away with it by accident, since the first
two bytes will still be correct, and the extra two bytes written (with
zeroes) will be overwritten correctly by the next store.
But for big-endian guests, this means the max_target field ends up as zero,
which means the guest will only recognize a single disk on the virtio-scsi
bus. This patch fixes the problem.
David Gibson [Fri, 23 Nov 2012 05:08:44 +0000 (16:08 +1100)]
virtio-scsi: Fix some endian bugs with virtio-scsi
The virtio-scsi specification does not specify the correct endianness for
fields in the request structure. It's therefore best to assume that it is
"guest native" endian since that's the (stupid and poorly defined) norm in
virtio.
However, the qemu device for virtio-scsi has no byteswaps at all, and so
will break if the guest has different endianness from the host. This patch
fixes it by adding tswap() calls for the sense_len and resid fields in
the request structure. In theory status_qualifier needs swaps as well,
but that field is never actually touched. The tag field is a uint64_t, but
since its value is completely arbitrary, it might as well be uint8_t[8]
and so it does not need swapping.
Peter Lieven [Sat, 17 Nov 2012 13:37:39 +0000 (14:37 +0100)]
iscsi: fix deadlock during login
If the connection is interrupted before the first login is successfully
completed qemu-kvm is waiting forever in qemu_aio_wait().
This is fixed by performing an sync login to the target. If the
connection breaks after the first successful login errors are
handled internally by libiscsi.
Michal Privoznik [Wed, 28 Nov 2012 10:46:39 +0000 (11:46 +0100)]
nbd-server-add: Fix the default for 'writable'
The documentation to this monitor command tells, that 'writable'
argument is optional and defaults to false. However, the code sets
true as the default. But since some applications may already been
using this, it's safer to fix the code and not documentation which
would break those applications.
Alex Horn [Mon, 26 Nov 2012 16:32:54 +0000 (17:32 +0100)]
rtc: Only call rtc_set_cmos when Register B SET flag is disabled.
This bug occurs when the SET flag of Register B is enabled. When an RTC
data register (i.e. any of the ten time/calender CMOS bytes) is set, the
data is (as expected) correctly stored in the cmos_data array. However,
since the SET flag is enabled, the function rtc_set_time is not invoked.
As a result, the field base_rtc in RTCState remains uninitialized. This
causes a problem on subsequent writes which can end up overwriting data.
To see this, consider writing data to Register A after having written
data to any of the RTC data registers; the following figure illustrates
the call stack for the Register A write operation:
Paolo Bonzini [Tue, 27 Nov 2012 08:16:24 +0000 (09:16 +0100)]
virtio-rng: do not use g_assert_cmpint
g_assert_cmpint is not available on glib 2.12, which is the minimum
version required to build QEMU (we only require 2.16 to run tests,
since that is the first version including GTester). Do not use it
in hardware models, use a normal assertion instead.
This fixes the buildbot failure for default_x86_64_rhel5.
Paolo Bonzini [Tue, 27 Nov 2012 08:51:48 +0000 (09:51 +0100)]
tests: make threadpool cancellation test looser
The cancellation test is failing on the buildbots. While the failure
merits a little more investigation to understand what is going on,
the logs show that the failure is not impacting the coverage
provided by the test. Hence, loosen a bit the assertions in a
way that should let the test proceed and hopefully pass.
Amit Shah [Wed, 21 Nov 2012 05:51:19 +0000 (11:21 +0530)]
virtio-rng: remove extra request for entropy
If we got fewer bytes from the backend than requested, don't poke the
backend for more bytes; the guest will ask for more (or if the guest has
already asked for more, the backend knows about it via handle_input()).
Amit Shah [Wed, 21 Nov 2012 05:51:18 +0000 (11:21 +0530)]
virtio-rng: use virtqueue_get_avail_bytes, fix migration
Popping an elem from the vq just to find out its length causes problems
with save/load later on. Use the new virtqueue_get_avail_bytes()
function instead, saves us the complexity in the migration code, as well
as makes the migration endian-safe.
Andreas Färber [Sun, 25 Nov 2012 19:44:33 +0000 (20:44 +0100)]
i8259: Fix PIC_COMMON() macro
It used a wrong struct type name since its introduction in 8f04ee0882aec9fe91fb70f767edf5dacff59835 (isa: pic: convert to QEMU
Object Model), apparently it is unused so far.
Eduardo Habkost [Thu, 22 Nov 2012 15:31:03 +0000 (13:31 -0200)]
target-i386: cpu: add missing flags to Haswell CPU model
When adding the Haswell CPU model, I intended to make it a superset of the
features present on the SandyBridge model, but I have removed the SEP and
RDTSCP features from the feature list by mistake. This patch adds the missing
SEP and RDTSCP features (that are present on SandyBridge) to Haswell.
Peter Maydell [Thu, 22 Nov 2012 16:48:45 +0000 (16:48 +0000)]
vl.c: Fix broken -usb option
Commit 094b287f0b accidentally broke the "-usb" command line
option, so it would have no effect if the user had not specified
any machine options at that point. (the return value from
'qemu_opts_find(qemu_find_opts("machine"), 0);' is NULL if there
are no user specified options, so it is only to be used for
looking up an option, not when trying to set one.) Similarly,
would '-usbdevice' no longer cause USB to default to enabled.
Fix this regression by using the same style of code for forcing
the usb=on machine option that we use for other aliases such as
'-enable-kvm'.
Paolo Bonzini [Fri, 23 Nov 2012 08:47:15 +0000 (09:47 +0100)]
qdev: simplify (de)allocation of buses
All conditional deallocation can now be done with object_delete.
Remove the @qom_allocated and @glib_allocated fields; replace the latter
with a direct assignment of the @free function pointer.
Paolo Bonzini [Fri, 23 Nov 2012 08:47:14 +0000 (09:47 +0100)]
qom: make object_delete usable for statically-allocated objects
Store in the object the freeing function that will be used at deletion
time. This makes it possible to use object_delete on statically-allocated
(embedded) objects. Dually, it makes it possible to use object_unparent
and object_unref without leaking memory, when the lifetime of object
might extend until after the call to object_delete.
Paolo Bonzini [Fri, 23 Nov 2012 08:47:13 +0000 (09:47 +0100)]
qdev: move bus removal to object_unparent
Add an ObjectClass method that is done at object_unparent time. It
should remove any backlinks to the object in the composition tree,
so that object_delete will be able to drop the last reference and
free the object.
Paolo Bonzini [Fri, 23 Nov 2012 08:47:12 +0000 (09:47 +0100)]
qom: fix refcount of non-heap-allocated objects
The reference count for embedded objects is always one too low, because
object_initialize_with_type returns with zero references to the object.
This causes premature finalization of the object (or an assertion failure)
after calling object_ref to add an extra reference and object_unref to
remove it.
The fix is to move the initial object_ref call from object_new_with_type
to object_initialize_with_type.
Heinz Graalfs [Mon, 19 Nov 2012 23:28:34 +0000 (23:28 +0000)]
s390: Fix ram_size updating in machine init
The global variable 'ram_size' is hidden by the local variable
declaration in s390_init(). Since we want to update the global
ram size in certain cases we must not use a local ram_size
variable.
- This fixes booting with unusual ram sizes like -m 67001
- This changes behaviour back to the situation before commit 5f072e1f3075bd869e0ace9f2545a85992ac0084
(create struct for machine initialization arguments)
Alexander Graf [Mon, 26 Nov 2012 18:49:58 +0000 (19:49 +0100)]
fbdev: fix pixman compile on old pixman
My QEMU compile failed with the following error:
qemu-pixman.c: In function ‘qemu_pixman_get_type’:
qemu-pixman.c:24: error: ‘PIXMAN_TYPE_BGRA’ undeclared (first use in this function)
qemu-pixman.c:24: error: (Each undeclared identifier is reported only once
qemu-pixman.c:24: error: for each function it appears in.)
Guard the PIXMAN_TYPE_BGRA branch like in the case right above the failing
case, so that compilation is fixed. Functionality on such old pixman is a
different question ;-).
Peter Maydell [Thu, 22 Nov 2012 06:48:45 +0000 (06:48 +0000)]
vl.c: Fix broken -usb option
Commit 094b287f0b accidentally broke the "-usb" command line
option, so it would have no effect if the user had not specified
any machine options at that point. (the return value from
'qemu_opts_find(qemu_find_opts("machine"), 0);' is NULL if there
are no user specified options, so it is only to be used for
looking up an option, not when trying to set one.) Similarly,
would '-usbdevice' no longer cause USB to default to enabled.
Fix this regression by using the same style of code for forcing
the usb=on machine option that we use for other aliases such as
'-enable-kvm'.
In one of the recent reworks to the XICS code, a bug was introduced where
we use the wrong sense and allocate level interrupts instead of message
interrupts for PCI MSIs. This patch fixes it.
kvm: fix incorrect length in a loop over kvm dirty pages map
QEMU allocates a map enough for 4k pages. However the system page size
can be 64K (for example on POWER) and the host kernel uses only a small
part of it as one big stores a dirty flag for 16 pages 4K each,
the hpratio variable stores this ratio and
the kvm_get_dirty_pages_log_range function handles it correctly.
However kvm_get_dirty_pages_log_range still goes beyond the data
provided by the host kernel which is not correct. It does not cause
errors at the moment as the whole bitmap is zeroed before doing KVM ioctl.
The patch reduces number of iterations over the map.
Julio Guerra [Fri, 19 Oct 2012 00:17:13 +0000 (00:17 +0000)]
PPC: Fix missing TRACE exception
This patch fixes bug 1031698 :
https://bugs.launchpad.net/qemu/+bug/1031698
If we look at the (truncated) translation of the conditional branch
instruction in the test submitted in the bug post, the call to the
exception helper is missing in the "bne-false" chunk of translated
code :
Peter Maydell [Tue, 20 Nov 2012 08:16:51 +0000 (08:16 +0000)]
hw/ide/macio: Fix segfault caused by NULL DMAContext*
Pass qemu_sglist_init the global dma_context_memory rather than a NULL
pointer; this fixes a segfault in dma_memory_map() when the guest
starts using DMA.
Eduardo Habkost [Thu, 22 Nov 2012 15:31:03 +0000 (13:31 -0200)]
target-i386/cpu: Add missing flags to Haswell CPU model
When adding the Haswell CPU model, I intended to make it a superset of the
features present on the SandyBridge model, but I have removed the SEP and
RDTSCP features from the feature list by mistake. This patch adds the
missing SEP and RDTSCP features (that are present on SandyBridge) to
Haswell.
Paolo Bonzini [Fri, 23 Nov 2012 15:56:18 +0000 (16:56 +0100)]
hmp: do not crash on invalid SCSI hotplug
Commit 0d93692 (qdev: Convert busses to QEMU Object Model, 2012-05-02)
removed a check on the type of the bus where a SCSI disk is hotplugged.
However, hot-plugging to the wrong kind of device now causes a crash
due to either a NULL pointer dereference (avoided by the previous patch)
or a failed QOM cast.
Instead, in this case we need to use object_dynamic_cast and check for
the result, similar to what was done before that commit.
Paolo Bonzini [Fri, 23 Nov 2012 14:59:43 +0000 (15:59 +0100)]
aio: avoid livelock behavior for Win32
The repeated calls to WaitForMultipleObjects may cause a livelock in aio_poll,
where no progress is made on bottom halves. This patch matches the behavior
of the POSIX code.
Isaku Yamahata [Wed, 14 Nov 2012 20:54:06 +0000 (15:54 -0500)]
q35: Introduce q35 pc based chipset emulator
pc q35 based chipset emulator to support pci express natively. Based on
Anthony Liguori's suggestion, the machine name is 'q35-next', with an alias
of 'q35'. At this point, there are no compatibility guarantees. When the
chipset stabilizes more, we will begin to version the machine names.
Major features which still need to be added:
-Migration support (mostly around ahci)
-ACPI hotplug support (pcie hotplug support is working)
-Passthrough support
Jason Baron [Wed, 14 Nov 2012 20:54:01 +0000 (15:54 -0500)]
pc: Move ioapic_init() from pc_piix.c to pc.c
Move ioapic_init() from pc_piix.c to pc.c, to make it a common function.
Rename ioapic_init() -> ioapic_init_gsi().
Move to pc.h so q35 can use them as well.
Anthony Liguori [Mon, 26 Nov 2012 14:19:20 +0000 (08:19 -0600)]
Merge remote-tracking branch 'kraxel/usb.72' into staging
* kraxel/usb.72:
usb-redir: Don't handle interrupt output packets async
usb-redir: Split usb_handle_interrupt_data into separate in/out functions
usb-smartcard-reader: Properly NAK interrupt eps when we've no events
usb-bt: Return NAK instead of STALL when interrupt ep has no data
uhci: Fix double unlink
uhci: Don't allow the guest to set port-enabled when there is no dev connected
uhci: Add a completions_only flag for async completions
Anthony Liguori [Mon, 26 Nov 2012 14:18:25 +0000 (08:18 -0600)]
Merge remote-tracking branch 'stefanha/block' into staging
* stefanha/block:
ide: Fix status register after short PRDs
ide: Fix crash with too long PRD
use int64_t for return values from rbd instead of int
vdi: don't override libuuid symbols
block: add bdrv_reopen() support for raw hdev, floppy, and cdrom
Aurelien Jarno [Sat, 24 Nov 2012 10:53:55 +0000 (11:53 +0100)]
tcg: mark local temps as MEM in dead_temp()
In dead_temp, local temps should always be marked as back to memory,
even if they have not been allocated (i.e. they are discared before
cross a basic block).
It fixes the following assertion in target-xtensa:
I check MIPS microMIPS manual [1], and found the major opcode might
be wrong. I add a comment to explicitly indicate what manual I am refering
to, and according that manual I remove microMIPS32 major opcodes 0x1f.
As for others, like 0x16, 0x17, 0x36 and 0x37, they are for higher-order
MIPS ISA level or new revision of this microMIPS architecture. Quote
from Johnson, they are belong MIPS64 [2].
Aurelien Jarno [Tue, 9 Oct 2012 19:53:11 +0000 (21:53 +0200)]
tcg/arm: fix cross-endian qemu_st16
The bswap16 TCG opcode assumes that the high bytes of the temp equal
to 0 before calling it. The ARM backend implementation takes this
assumption to slightly optimize the generated code.
The same implementation is called for implementing the cross-endian
qemu_st16 opcode, where this assumption is not true anymore. One way to
fix that would be to zero the high bytes before calling it. Given the
store instruction just ignore them, it is possible to provide a slightly
more optimized version. With ARMv6+ the rev16 instruction does the work
correctly. For lower ARM versions the patch provides a version which
behaves correctly with non-zero high bytes, but fill them with junk.
Aurelien Jarno [Tue, 9 Oct 2012 19:53:11 +0000 (21:53 +0200)]
tcg/arm: fix TLB access in qemu-ld/st ops
The TCG arm backend considers likely that the offset to the TLB
entries does not exceed 12 bits for mem_index = 0. In practice this is
not true for at least the MIPS target.
The current patch fixes that by loading the bits 23-12 with a separate
instruction, and using loads with address writeback, independently of
the value of mem_idx. In total this allow a 24-bit offset, which is a
lot more than needed.
Instead report them as successfully completed directly on submission, this
has 2 advantages:
1) This matches the timing of interrupt output packets on real hardware,
with the previous async handling, if an ep has an interval of say 500 ms,
then there would be 500+ ms between the submission and the guest seeing the
completion, as we wont do the write back until the qh gets polled again. And
in the mean time the guest may very well have timed out, as the guest can
reasonable expect a much quicker completion.
2) This fixes interrupt output packets potentially getting send twice
surrounding a migration. As we delay the writeback to guest memory until
the qh gets polled again, there is a window between completion and writeback
where migration can happen, in this case the destination will not know
about the completion, and it will execute the packet *again*
But it does also come with a disadvantage:
1) If the actual interrupt out to the real usb device fails, there is no
way to report this back to the guest.
This patch assumes however that interrupt outs in practice never fail, as
they are only used by specialized drivers, which are unlikely to issue illegal
requests (unlike general class drivers which often issue requests which some
devices don't implement). And that thus the advantages outway the disadvantage.
Hans de Goede [Sat, 17 Nov 2012 11:11:50 +0000 (12:11 +0100)]
uhci: Don't allow the guest to set port-enabled when there is no dev connected
It is possible for device disconnect and the guest trying to reset the port
(because of USB xact errors prior to the disconnect getting signaled) to race,
when we hit this race, the guest will write the port-control register with its
pre-disconnect value + the reset bit set, after which we have a disconnected
device with its port-enabled bit set in its port-control register, which
is no good :)
Hans de Goede [Sat, 17 Nov 2012 11:11:49 +0000 (12:11 +0100)]
uhci: Add a completions_only flag for async completions
Add a completions_only flag, and set this when running process_frame for async
completion handling, this fixes 2 issues in a single patch:
1) It makes sure async completed packets get written to guest mem immediately,
even if all the bandwidth for the frame was consumed from the timer run
process_frame. This is necessary as delaying their writeback to the next frame
can cause the completion to get lost on migration.
2) The calling of process_frame from a bh on async completion causes iso
tds to get server more often they should, messing up usb sound class device
timing. By only processing completed packets, the iso tds get skipped fixing
this.
Lei Li [Mon, 19 Nov 2012 09:15:08 +0000 (17:15 +0800)]
Fix the inconsistency in x509-dh-key-file parameter
Commit c448e8552b0f6135dabddf944a7110f929c08320 (spice: tls
support) added options to the -spice command line. But there
is an inconsistency between the declaration of the option
'x509-dh-key-file' to -spice command line and its parameter
parsing 'x509-dh-file' in function qemu_spice_init.
Kevin Wolf [Tue, 20 Nov 2012 16:27:44 +0000 (17:27 +0100)]
ide: Fix status register after short PRDs
When failing a request because the length of the regions described by
the PRDT was too short for the requested number of sectors, the IDE
emulation forgot to update the status register, so that the device would
keep the BSY flag set indefinitely.
Stefan Priebe [Tue, 20 Nov 2012 12:44:55 +0000 (13:44 +0100)]
use int64_t for return values from rbd instead of int
rbd / rados tends to return pretty often length of writes
or discarded blocks. These values might be bigger than int.
The steps to reproduce are:
mkfs.xfs -f a whole device bigger than int in bytes. mkfs.xfs sends
a discard. Important is that you use scsi-hd and set
discard_granularity=512. Otherwise rbd disabled discard support.
Stefan Hajnoczi [Tue, 20 Nov 2012 15:34:17 +0000 (16:34 +0100)]
vdi: don't override libuuid symbols
It's poor symbol hygiene to provide a global symbols that collide with a
common library like libuuid. If QEMU links against a shared library
that depends on uuid_generate() it can end up calling our stub version
of the function.
This exact scenario happened with GlusterFS libgfapi.so, which depends
on libglusterfs.so's uuid_generate().
Scope the uuid stubs for vdi.c only and avoid affecting other shared
objects.
Jeff Cody [Tue, 20 Nov 2012 15:21:10 +0000 (10:21 -0500)]
block: add bdrv_reopen() support for raw hdev, floppy, and cdrom
For hdev, floppy, and cdrom, the reopen() handlers are the same as
for the file reopen handler. For floppy and cdrom types, however,
we keep O_NONBLOCK, as in the _open function.
For tap, we currently assume the vnet header size is 10
(the default value) but that might not be the case
if tap is persistent and has been used by qemu previously.
To fix, set host header size in tap device on open.
Anthony Liguori [Mon, 19 Nov 2012 15:27:13 +0000 (09:27 -0600)]
Merge remote-tracking branch 'bonzini/build-urgent' into staging
* bonzini/build-urgent:
Makefile: Add missing dependency (fix parallel builds)
tests: link in stubs
libcacard: link in stubs
libcacard: make unnesting rules available to Makefile.objs
projects / qemu.git / log