qemu
15 years agobuild system: silent generation of doc files and qemu-options.h (Jan Kiszka)
aliguori [Sun, 5 Apr 2009 17:40:46 +0000 (17:40 +0000)]
build system: silent generation of doc files and qemu-options.h (Jan Kiszka)

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/trunk@6979 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAdd host_device support to qemu-img. (Nolan Leake)
aliguori [Sun, 5 Apr 2009 17:40:43 +0000 (17:40 +0000)]
Add host_device support to qemu-img. (Nolan Leake)

This patch allows the use a host_device as the destination for "qemu-img
convert".

I added a ->bdrv_create function host_device.  It merely verifies that
the device exists and is large enough.

A check is needed in the qemu-img convert loop to ensure that we write
out all 0 sectors to the host_device.  Otherwise they end up with stale
garbage where all zero sectors were expected.

I also made the check against bdrv_is_allocated enabled for everything
_except_ host devices, since there is no point in making the block
backend write a bunch of zeros just so that we can memcmp them
immediately afterwards.  Host devices can't benefit from this because
there is no way to differentiate between a sector being unallocated
because it was never written, or because it was written with all zeros
and then made a trip through qemu-img convert.

Finally, there is an unrelated fix for a typo in the error message
printed if the destination device does not support ->bdrv_create.

Signed-off-by: Nolan Leake <nolan <at> sigbus.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6978 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix (at least one cause of) qcow2 corruption. (Nolan Leake)
aliguori [Sun, 5 Apr 2009 17:40:38 +0000 (17:40 +0000)]
Fix (at least one cause of) qcow2 corruption. (Nolan Leake)

qcow2's get_cluster_offset() scans forward in the l2 table to find other
clusters that have the same allocation status as the first cluster.
This is used by (among others) qcow_is_allocated().

Unfortunately, it was not checking to be sure that it didn't fall off
the end of the l2 table.  This patch adds that check.

The symptom that motivated me to look into this was that
bdrv_is_allocated() was returning false when there was in fact data
there.  This is one of many ways this bug could lead to data corruption.

I checked the other place that scans for consecutive unallocated blocks
(alloc_cluster_offset()) and it appears to be OK:
    nb_clusters = MIN(nb_clusters, s->l2_size - l2_index);
appears to prevent the same problem from occurring.

Signed-off-by: Nolan Leake <nolan <at> sigbus.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6977 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDocument QEMU coding style (v2) (Avi Kivity)
aliguori [Sun, 5 Apr 2009 17:40:34 +0000 (17:40 +0000)]
Document QEMU coding style (v2) (Avi Kivity)

With the help of some Limoncino I noted several aspects of the QEMU coding
style, particularly where it differs from the Linux coding style as many
contributors work on both projects.

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@6976 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix oops on 2.6.25 guest (Rusty Russell)
aliguori [Sun, 5 Apr 2009 17:40:08 +0000 (17:40 +0000)]
Fix oops on 2.6.25 guest (Rusty Russell)

I believe this is behind the following:
https://bugs.edge.launchpad.net/ubuntu/jaunty/+source/linux/+bug/331128

virtio_pci in 2.6.25 didn't do feature negotiation correctly: it acked every
bit.  Fortunately, we can detect this.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6975 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix branches and TLB matches for 64 bit targets
blueswir1 [Sat, 4 Apr 2009 19:10:26 +0000 (19:10 +0000)]
Fix branches and TLB matches for 64 bit targets

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6974 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAllocate space for static call args, increase stack frame size on Sparc64
blueswir1 [Sat, 4 Apr 2009 15:33:03 +0000 (15:33 +0000)]
Allocate space for static call args, increase stack frame size on Sparc64

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6973 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoCompile all files with -ffixed-g5 etc. to avoid env (%g5) corruption
blueswir1 [Sat, 4 Apr 2009 09:21:28 +0000 (09:21 +0000)]
Compile all files with -ffixed-g5 etc. to avoid env (%g5) corruption

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6972 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoMake i386-softmmu boot on Sparc host
blueswir1 [Sat, 4 Apr 2009 07:41:20 +0000 (07:41 +0000)]
Make i386-softmmu boot on Sparc host

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6971 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH: Fix linux-user _is_cached typo.
edgar_igl [Fri, 3 Apr 2009 07:33:24 +0000 (07:33 +0000)]
SH: Fix linux-user _is_cached typo.

Signed-off-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6970 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH: Add cpu_sh4_is_cached for linux-user.
edgar_igl [Fri, 3 Apr 2009 07:29:38 +0000 (07:29 +0000)]
SH: Add cpu_sh4_is_cached for linux-user.

The entire U0 area is assumed to be cacheable.

Signed-off-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6969 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoImmediate versions of ro[lr]
malc [Thu, 2 Apr 2009 22:54:35 +0000 (22:54 +0000)]
Immediate versions of ro[lr]

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6968 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoTemporary workaround for ppc on ppc
malc [Thu, 2 Apr 2009 01:16:39 +0000 (01:16 +0000)]
Temporary workaround for ppc on ppc

target-ppc/translate.c puts values of type opcode_t into .opcodes
section, using GCC extension to do so, and hoping that this will make
them appear contiguously and in the source order in the resulting
executable. This assumption is not safe and is known to be violated
with certain versions of GCC, certain flags passed to it and on
certain platforms (gcc 4.3.0, -O and PPC/PPC64 for instance)

The workaround consists of adding -fno-unit-at-a-time to the list of
GCC command line options while building PPC translate.o on a PPC.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6967 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH: Improve movca.l/ocbi emulation.
edgar_igl [Wed, 1 Apr 2009 23:10:46 +0000 (23:10 +0000)]
SH: Improve movca.l/ocbi emulation.

Author: Vladimir Prus <vladimir@codesourcery.com>

    Fix movcal.l/ocbi emulation.

        * target-sh4/cpu.h (memory_content): New.
        (CPUSH4State): New fields movcal_backup and movcal_backup_tail.
        * target-sh4/helper.h (helper_movcal)
        (helper_discard_movcal_backup, helper_ocbi): New.
        * target-sh4/op_helper.c (helper_movcal)
        (helper_discard_movcal_backup, helper_ocbi): New.
        * target-sh4/translate.c (DisasContext): New field has_movcal.
        (sh4_defs): Update CVS for SH7785.
        (cpu_sh4_init): Initialize env->movcal_backup_tail.
        (_decode_opc): Discard movca.l-backup.
        Make use of helper_movcal and helper_ocbi.
        (gen_intermediate_code_internal): Initialize has_movcal to 1.

Thanks to Shin-ichiro KAWASAKI and Paul Mundt for valuable feedback.

Acked-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6966 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoImplement and use shared memory framebuffer device rendering reoutine.
pbrook [Wed, 1 Apr 2009 12:27:59 +0000 (12:27 +0000)]
Implement and use shared memory framebuffer device rendering reoutine.
Use DMA mapping API.

Signed-off-by: Paul Brook <paul@codesourcery.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6965 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoUse pixel_ops.h
pbrook [Wed, 1 Apr 2009 11:43:02 +0000 (11:43 +0000)]
Use pixel_ops.h

Signed-off-by: Paul Brook <paul@codesourcery.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6964 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoRemove unused variable
blueswir1 [Tue, 31 Mar 2009 17:18:10 +0000 (17:18 +0000)]
Remove unused variable

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6963 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoUpdate OpenBIOS images to r482
blueswir1 [Tue, 31 Mar 2009 15:46:09 +0000 (15:46 +0000)]
Update OpenBIOS images to r482

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6962 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAvoid rounding problems in ptimer_get_count
pbrook [Tue, 31 Mar 2009 14:34:24 +0000 (14:34 +0000)]
Avoid rounding problems in ptimer_get_count

Signed-off-by: Paul Brook <paul@codesourcery.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6961 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix wrong return value
blueswir1 [Mon, 30 Mar 2009 17:51:29 +0000 (17:51 +0000)]
Fix wrong return value

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6959 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoRemove dead AIO code for win32
aliguori [Sun, 29 Mar 2009 20:54:17 +0000 (20:54 +0000)]
Remove dead AIO code for win32

AFAIK, it's never been enabled and hasn't even compiled in ages.  Remove the
code to avoid confusion.  If someone ever decides to fix it, they can get at
the old version in RCS history.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6958 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_movcf_*()
aurel32 [Sun, 29 Mar 2009 15:45:02 +0000 (15:45 +0000)]
target-mips: optimize gen_movcf_*()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6957 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_movci()
aurel32 [Sun, 29 Mar 2009 15:44:50 +0000 (15:44 +0000)]
target-mips: optimize gen_movci()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6956 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_compute_branch1()
aurel32 [Sun, 29 Mar 2009 15:44:34 +0000 (15:44 +0000)]
target-mips: optimize gen_compute_branch1()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6955 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoMisc scsi disk/cdrom fixes/improvements 4/4
blueswir1 [Sun, 29 Mar 2009 15:40:59 +0000 (15:40 +0000)]
Misc scsi disk/cdrom fixes/improvements 4/4

Implement Test Unit Ready command (return NOT READY as above
if !bdrv_is_inserted(...))

Signed-off-by: Juergen Lock <nox@jelal.kn-bremen.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6954 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agomisc scsi disk/cdrom fixes/improvements 3/4
blueswir1 [Sun, 29 Mar 2009 15:40:42 +0000 (15:40 +0000)]
misc scsi disk/cdrom fixes/improvements 3/4

Add asc 0x3a, ascq 0: Medium not present to NOT READY sense
(needed to keep some guests from retrying causing long sleeps in the
kernel)

Signed-off-by: Juergen Lock <nox@jelal.kn-bremen.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6953 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agomisc scsi disk/cdrom fixes/improvements 2/4
blueswir1 [Sun, 29 Mar 2009 15:40:21 +0000 (15:40 +0000)]
misc scsi disk/cdrom fixes/improvements 2/4

Implement cdrom load/eject functionality (Start Stop Unit command)

Signed-off-by: Juergen Lock <nox@jelal.kn-bremen.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6952 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agomisc scsi disk/cdrom fixes/improvements 1/4
blueswir1 [Sun, 29 Mar 2009 15:39:59 +0000 (15:39 +0000)]
misc scsi disk/cdrom fixes/improvements 1/4

Use correct sector size for cdrom Read TOC command

Signed-off-by: Juergen Lock <nox@jelal.kn-bremen.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6951 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: don't map FP registers as TCG global variables
aurel32 [Sun, 29 Mar 2009 15:39:08 +0000 (15:39 +0000)]
target-mips: don't map FP registers as TCG global variables

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6950 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: fix divu instruction
aurel32 [Sun, 29 Mar 2009 15:36:47 +0000 (15:36 +0000)]
target-mips: fix divu instruction

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6949 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotcg: fix _tl aliases for divu/remu
aurel32 [Sun, 29 Mar 2009 14:08:54 +0000 (14:08 +0000)]
tcg: fix _tl aliases for divu/remu

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6948 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-ppc: Explain why the whole TLB is flushed on SR write
aurel32 [Sun, 29 Mar 2009 13:36:32 +0000 (13:36 +0000)]
target-ppc: Explain why the whole TLB is flushed on SR write

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6947 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix hxtool eating backslash sequences for sh != bash
blueswir1 [Sun, 29 Mar 2009 10:50:43 +0000 (10:50 +0000)]
Fix hxtool eating backslash sequences for sh != bash

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6946 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix hxtool eating backslash sequences
blueswir1 [Sun, 29 Mar 2009 09:06:43 +0000 (09:06 +0000)]
Fix hxtool eating backslash sequences

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6945 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agofix format string warnings in block-qcow2.c (Christoph Hellwig)
aliguori [Sun, 29 Mar 2009 01:31:56 +0000 (01:31 +0000)]
fix format string warnings in block-qcow2.c (Christoph Hellwig)

Recent patches added two compiler warnings about the format string
usage in qcow_read_extensions.  One is printing a uint64_t using
%lu which is incorrect on many platforms as it can be a unsigned
long long, the second one is printing the result of sizeof as
%lu, but it is a size_t so it needs to be printed using %zu.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6944 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agocheck for bs->drv in bdrv_flush (Christoph Hellwig)
aliguori [Sun, 29 Mar 2009 01:31:51 +0000 (01:31 +0000)]
check for bs->drv in bdrv_flush (Christoph Hellwig)

All the bdrv_ helpers should check for bs->drv being zero as that means
there is no backend image open.  bdrv_flush fails to perform that check
and can thus cause NULL pointer dereferences.

Found using qemu-io.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6943 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoremove dead code in bdrv_check_request (Christoph Hellwig)
aliguori [Sun, 29 Mar 2009 01:31:48 +0000 (01:31 +0000)]
remove dead code in bdrv_check_request (Christoph Hellwig)

Remove code dealing with negative sector numbers for byte access in
bdrv_check_request as sector numbers can't ever be negative.

Previously we supported negative sector counts for byte access, but
never sector numbers.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6942 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize write to env->hflags
aurel32 [Sun, 29 Mar 2009 01:28:29 +0000 (01:28 +0000)]
target-mips: optimize write to env->hflags

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6941 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_muldiv()
aurel32 [Sun, 29 Mar 2009 01:19:31 +0000 (01:19 +0000)]
target-mips: optimize gen_muldiv()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6940 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotcg: add _tl aliases for div/divu/rem/remu
aurel32 [Sun, 29 Mar 2009 01:19:22 +0000 (01:19 +0000)]
tcg: add _tl aliases for div/divu/rem/remu

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6939 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_HILO()
aurel32 [Sun, 29 Mar 2009 01:19:12 +0000 (01:19 +0000)]
target-mips: optimize gen_HILO()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6938 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_trap()
aurel32 [Sun, 29 Mar 2009 01:19:02 +0000 (01:19 +0000)]
target-mips: optimize gen_trap()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6937 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_compute_branch()
aurel32 [Sun, 29 Mar 2009 01:18:52 +0000 (01:18 +0000)]
target-mips: optimize gen_compute_branch()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6936 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: don't mix result and arguments in gen_op_*
aurel32 [Sun, 29 Mar 2009 01:18:43 +0000 (01:18 +0000)]
target-mips: don't mix result and arguments in gen_op_*

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6935 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: gen_bshfl()
aurel32 [Sun, 29 Mar 2009 01:18:34 +0000 (01:18 +0000)]
target-mips: gen_bshfl()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6934 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_mul_vr54xx()
aurel32 [Sun, 29 Mar 2009 01:18:24 +0000 (01:18 +0000)]
target-mips: optimize gen_mul_vr54xx()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6933 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: optimize gen_cl()
aurel32 [Sun, 29 Mar 2009 01:18:16 +0000 (01:18 +0000)]
target-mips: optimize gen_cl()

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6932 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-ppc: avoid nop to override next instruction
aurel32 [Sun, 29 Mar 2009 01:18:03 +0000 (01:18 +0000)]
target-ppc: avoid nop to override next instruction

While searching PC, always store the pc of a new instruction.
Instructions that didn't generate tcg code (such as nop) prevented the
next one to be referenced.

Based on patch for target-alpha, r6930.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6931 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: bug fix: avoid nop to override next instruction
aurel32 [Sun, 29 Mar 2009 01:04:39 +0000 (01:04 +0000)]
target-alpha: bug fix: avoid nop to override next instruction

While searching PC, always store the pc of a new instruction.
Instructions that didn't generate tcg code (such as nop) prevented the next
one to be referenced.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6930 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: remove PALCODE_ declarations (unused).
aurel32 [Sun, 29 Mar 2009 00:22:12 +0000 (00:22 +0000)]
target-alpha: remove PALCODE_ declarations (unused).

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6929 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: add instruction name in comments for hw_ld opcode.
aurel32 [Sun, 29 Mar 2009 00:22:01 +0000 (00:22 +0000)]
target-alpha: add instruction name in comments for hw_ld opcode.

Make code slightly easier to read.
Also unused hw_ld opcodes now generate an invalid opc fault.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6928 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDocument which IPR are used by 21264
aurel32 [Sun, 29 Mar 2009 00:21:43 +0000 (00:21 +0000)]
Document which IPR are used by 21264

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6927 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: tb_flush helper should flush the tb (and not the tlb).
aurel32 [Sun, 29 Mar 2009 00:14:16 +0000 (00:14 +0000)]
target-alpha: tb_flush helper should flush the tb (and not the tlb).

Looks to be a typo fix.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6926 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: fix temp free for hw_st
aurel32 [Sun, 29 Mar 2009 00:14:06 +0000 (00:14 +0000)]
target-alpha: fix temp free for hw_st

No need to stop translation after hw_st.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6925 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: fix bug: integer conditional branch offset is 21 bits wide.
aurel32 [Sun, 29 Mar 2009 00:13:56 +0000 (00:13 +0000)]
target-alpha: fix bug: integer conditional branch offset is 21 bits wide.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6924 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: Fix bug: do not mask address LSBs for ldwu.
aurel32 [Sun, 29 Mar 2009 00:13:47 +0000 (00:13 +0000)]
target-alpha: Fix bug: do not mask address LSBs for ldwu.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6923 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-alpha: Fix bug: palcode is at least 6 bits.
aurel32 [Sun, 29 Mar 2009 00:13:38 +0000 (00:13 +0000)]
target-alpha: Fix bug: palcode is at least 6 bits.

Signed-off-by: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6922 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agobuild system: silence gdbstub-xml.c generation
aurel32 [Sat, 28 Mar 2009 23:46:00 +0000 (23:46 +0000)]
build system: silence gdbstub-xml.c generation

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6921 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotcg/README: fix description of bswap32_i32/i64
aurel32 [Sat, 28 Mar 2009 23:27:30 +0000 (23:27 +0000)]
tcg/README: fix description of bswap32_i32/i64

Thanks to Stuart Brady for the notice.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6920 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-sh4: r2d --append option support
aurel32 [Sat, 28 Mar 2009 23:18:47 +0000 (23:18 +0000)]
target-sh4: r2d --append option support

Remove qemu_ram_alloc(SDRAM_BASE) and related changes.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6919 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agohw/eeprom93xx.c: support 93xx EEPROMs with more than 255 words
aurel32 [Sat, 28 Mar 2009 23:14:52 +0000 (23:14 +0000)]
hw/eeprom93xx.c: support 93xx EEPROMs with more than 255 words

In the head of eeprom93xx.c we promise to support chips with 256 words,
but store the size in an unsigned byte. This patch replaces this with an
16 bit variable and changes the load/store code accordingly (introducing a
new version).

Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6918 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agohw/eeprom93xx.c: substitute structure dump with discrete dump in eeprom_save/load
aurel32 [Sat, 28 Mar 2009 23:14:41 +0000 (23:14 +0000)]
hw/eeprom93xx.c: substitute structure dump with discrete dump in eeprom_save/load

The EEPROM 93xx device used to dump a C structure to the migration stream.
This structure includes mixed 8 and 16bit variables and is thus subject to
compiler dependent padding. Replace this with discrete dumps of each member
(and add a padding byte to ensure compatibility, a version update is
included in the following patch).

Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6917 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-sh4: r2d --append option support
aurel32 [Sat, 28 Mar 2009 23:14:32 +0000 (23:14 +0000)]
target-sh4: r2d --append option support

Add linux kernel command line ("--append" option) support.
Fix kernel loading address to appropriate position when --append used.
Using --kernel but --append case is left untouched for backward compatibility.

This also change the host<->SH address mapping for r2d to
 host addr == phys_ram_base + SH addr.

Signed-off-by: Takashi YOSHII <takasi-y@ops.dti.ne.jp>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6916 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: fix FPU in 64-bit mode
aurel32 [Sat, 28 Mar 2009 22:22:50 +0000 (22:22 +0000)]
target-mips: fix FPU in 64-bit mode

TCG does not allow the same memory location to be aliased in two
different global registers, fpu_fpr32 and fpu_fpr64.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6915 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-mips: implement FPU Flush-To-Zero mode
aurel32 [Sat, 28 Mar 2009 22:22:40 +0000 (22:22 +0000)]
target-mips: implement FPU Flush-To-Zero mode

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6914 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agogdbstub: Allow re-instantiation (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 18:05:53 +0000 (18:05 +0000)]
gdbstub: Allow re-instantiation (Jan Kiszka)

[ Note: depends on char closing fixes ]

Properly clean up the gdbstub when the user tries to re-open it
(possibly under a different address). Moreover, allow to shut it down
from the monitor via 'gdbserver none'.

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/trunk@6913 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agochar: Fix closing of various char devices (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 17:58:14 +0000 (17:58 +0000)]
char: Fix closing of various char devices (Jan Kiszka)

This patch fixes several issues around closing char devices. Affected
were pty (timer was left behind, even running), udp (no close handling
at all) and tcp (missing async IO handler cleanup). The bugs either
caused segfaults or stalled the qemu process. So far, hot-unplugging USB
serial adapters suffered from this.

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/trunk@6911 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoqemu-img: adding a "-F base_fmt" option to "qemu-img create -b" (Uri Lublin)
aliguori [Sat, 28 Mar 2009 17:55:19 +0000 (17:55 +0000)]
qemu-img: adding a "-F base_fmt" option to "qemu-img create -b" (Uri Lublin)

If the user specifies the backing file format,
then when opening the backing file, there is no need
to probe the (backing file) image to figure out its format.

This follows my previous patches implementing bdrv_create2
which keeps (for qcow2 only) the backing file format
as a qcow2-extension

Suggested by Daniel P. Berrange.

Signed-off-by: Uri Lublin <uril@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6910 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoblock-qcow2: keep backing file format in a qcow2 extension (Uri Lublin)
aliguori [Sat, 28 Mar 2009 17:55:14 +0000 (17:55 +0000)]
block-qcow2: keep backing file format in a qcow2 extension (Uri Lublin)

Use a qcow2 extension to keep the backing file format.

By keeping the backing file format, we can:
1. Provide a way to know the backing file format without probing
   it (setting the format at creation time).
2. Enable using qcow2 format over host block devices.
   (only if the user specifically asks for it, by providing the format
   at creation time).

Also fixes a security flaw found by Daniel P. Berrange on [1]
which summarizes: "Autoprobing: just say no."

[1] http://lists.gnu.org/archive/html/qemu-devel/2008-12/msg01083.html

Signed-off-by: Uri Lublin <uril@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6909 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoblock: support known backing format for image create and open (Uri Lublin)
aliguori [Sat, 28 Mar 2009 17:55:10 +0000 (17:55 +0000)]
block: support known backing format for image create and open (Uri Lublin)

Added a backing_format field to BlockDriverState.
Added bdrv_create2 and drv->bdrv_create2 to create an image with
a known backing file format.
Upon bdrv_open2 if backing format is known use it, instead of
probing the (backing) image.

Signed-off-by: Uri Lublin <uril@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6908 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoIntroducing qcow2 extensions (Uri Lublin)
aliguori [Sat, 28 Mar 2009 17:55:06 +0000 (17:55 +0000)]
Introducing qcow2 extensions (Uri Lublin)

Qcow2 extensions are build of magic (id) len (in bytes) and data.
They reside right after the qcow2 header.
If a backing filename exists it follows the qcow2 extension (if exist)

Qcow2 extensions are read upon image open.
Qcow2 extensions are identified by their magic.
Unknown qcow2 extensions (unknown magic) are skipped.
A Special magic of 0 means end-of-qcow2-extensions.

In this patchset, to be used to keep backing file format.

Based on a work done by Shahar Frank <sfrank@redhat.com>.

Signed-off-by: Uri Lublin <uril@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6907 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agokvm: Drop kvm_patch_opcode_byte (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 17:51:40 +0000 (17:51 +0000)]
kvm: Drop kvm_patch_opcode_byte (Jan Kiszka)

As cpu_memory_rw_debug is now capable of modifying ROM, we can drop our
own patch function.

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/trunk@6906 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoROM write access for debugging (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 17:51:36 +0000 (17:51 +0000)]
ROM write access for debugging (Jan Kiszka)

Enhance cpu_memory_rw_debug so that it can write even to ROM regions.
This allows to modify ROM via gdb (I see no point in denying this to the
user), and it will enable us to drop kvm_patch_opcode_byte().

Credits go to Avi for suggesting this.

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/trunk@6905 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoUse the DMA api to map virtio elements.
aliguori [Sat, 28 Mar 2009 17:46:18 +0000 (17:46 +0000)]
Use the DMA api to map virtio elements.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6904 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agovirtio-blk: use generic vectored I/O APIs (Christoph Hellwig)
aliguori [Sat, 28 Mar 2009 17:46:14 +0000 (17:46 +0000)]
virtio-blk: use generic vectored I/O APIs (Christoph Hellwig)

Use the generic bdrv_aio_readv/bdrv_aio_writev APIs instead of linearizing
buffers directly.  This enables using the future native preadv/pwritev
support.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6903 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoadd qemu_iovec_init_external (Christoph Hellwig)
aliguori [Sat, 28 Mar 2009 17:46:10 +0000 (17:46 +0000)]
add qemu_iovec_init_external (Christoph Hellwig)

Allow to initialize a QEMUIOVector from an externally allocated iovec.
qiov->nalloc is initialized to -1 to indicate external storage for qiov->iov
and all functions dealing with memory management assert on the iovec beeing
an internally managed first.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6902 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoClean some PCI defines (Stefan Weil)
aliguori [Sat, 28 Mar 2009 17:29:07 +0000 (17:29 +0000)]
Clean some PCI defines (Stefan Weil)

this patch adds some more defines from linux/pci_regs.h to
hw/pci.h. There is now no longer a need to define them in
eepro100.c, so they were removed there.

Some defines from linux/pci_regs.h had similar, but not
the same defines in hw/pci.h (PCI_REVISION_ID / PCI_REVISION,
PCI_SUBSYSTEM_VENDOR_ID / PCI_SUBVENDOR_ID,
PCI_SUBSYSTEM_ID / PCI_SUBDEVICE_ID).

I suggest to use the "standard" from linux/pci_regs.h and
replace the "old" Qemu ones. To facilitate the migration,
my patch does not remove the old defines but marks them
as obsolete. After a migration to the "standard" defines,
pci.h could use linux/pci_regs.h which is far more complete.

The patch is needed for an updated maintainer version of
hw/eepro100.c which I'd like to see in Qemu stable.

 * Remove declarations already declared in header file from eepro100.c
 * Add missing declarations from pci_regs.h to pci.h
 * Mark "non-standard" declarations in pci.h as obsolete

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6901 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix monitor command (screendump) (Stefan Weil)
aliguori [Sat, 28 Mar 2009 17:29:02 +0000 (17:29 +0000)]
Fix monitor command (screendump) (Stefan Weil)

starting with r6839, the monitor command 'screendump'
raises a nullpointer memory access which crashs Qemu.

Fix crash when calling screendump from monitor.

This was a regression introduced with r6839:
DisplayAllocator interface (Stefano Stabellini)

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6900 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoRemove nodisk_ok machine feature (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 17:28:58 +0000 (17:28 +0000)]
Remove nodisk_ok machine feature (Jan Kiszka)

All archs have some kind of firmware to load and can be fine with it
already. So there is not much use in enforcing the presence of a disk.
If the system setup requires one, the user will notice it anyway once
the firmware/bios fails to boot from it.

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/trunk@6899 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agomusicpal: Fix regression caused by 6839 (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 17:28:53 +0000 (17:28 +0000)]
musicpal: Fix regression caused by 6839 (Jan Kiszka)

Pass correct DisplayState field to is_surface_bgr().

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6898 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agogdbstub: Drop redundant memset after qemu_mallocz (Jan Kiszka)
aliguori [Sat, 28 Mar 2009 17:28:49 +0000 (17:28 +0000)]
gdbstub: Drop redundant memset after qemu_mallocz (Jan Kiszka)

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/trunk@6897 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoget roms more room. (Glauber Costa)
aliguori [Sat, 28 Mar 2009 17:28:45 +0000 (17:28 +0000)]
get roms more room. (Glauber Costa)

This patch increases by 50 % the size available for option roms.
The main motivator is that some roms grew bigger than the 64k we
currently allocate for them (Hey, it's 2009!)

One example is the gpxe project, that produces some roms with 69k,
70k, etc. The space proposed by this patch actually makes it as
big as 84k. Probably still a fit for some time.

But there is no free lunch. This space must come from somewhere,
and we take it from vga rom space. Currently, our vga roms are
around 35k in size. With this patch, option rom space will begin
just after vga ends, aligned to the next 2k boundary.

Technicaly, we could do the same with the uper space (the bios itself),
but since bochs bios is already 128 k in size, I don't see an
urgent need to do it.

[ fix case for vgabioses smaller than 30k, by Carl-Daniel Hailfinger ]

Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6896 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agonew scsi-generic abstraction, use SG_IO (Christoph Hellwig)
aliguori [Sat, 28 Mar 2009 17:28:41 +0000 (17:28 +0000)]
new scsi-generic abstraction, use SG_IO (Christoph Hellwig)

Okay, I started looking into how to handle scsi-generic I/O in the
new world order.

I think the best is to use the SG_IO ioctl instead of the read/write
interface as that allows us to support scsi passthrough on disk/cdrom
devices, too.  See Hannes patch on the kvm list from August for an
example.

Now that we always do ioctls we don't need another abstraction than
bdrv_ioctl for the synchronous requests for now, and for asynchronous
requests I've added a aio_ioctl abstraction keeping it simple.

Long-term we might want to move the ops to a higher-level abstraction
and let the low-level code fill out the request header, but I'm lazy
enough to leave that to the people trying to support scsi-passthrough
on a non-Linux OS.

Tested lightly by issuing various sg_ commands from sg3-utils in a guest
to a host CDROM device.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6895 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDocument sun ID PROM contents
blueswir1 [Sat, 28 Mar 2009 16:17:50 +0000 (16:17 +0000)]
Document sun ID PROM contents

Signed-off-by: Robert Reif <reif@earthlink.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6894 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix DMA API when handling an immediate error from block layer (Avi Kivity)
aliguori [Sat, 28 Mar 2009 16:11:25 +0000 (16:11 +0000)]
Fix DMA API when handling an immediate error from block layer (Avi Kivity)

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

15 years agoFix vectored aio bounce handling immediate errors (Avi Kivity)
aliguori [Sat, 28 Mar 2009 16:11:20 +0000 (16:11 +0000)]
Fix vectored aio bounce handling immediate errors (Avi Kivity)

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

15 years agohost_device_remove: remove incorrect check for device name (Eduardo Habkost)
aliguori [Sat, 28 Mar 2009 15:58:58 +0000 (15:58 +0000)]
host_device_remove: remove incorrect check for device name (Eduardo Habkost)

There is no need to check for valid prefixes on the the device name
when removing it. If the device name is found on the vlan client list,
it can be removed, regardless of the prefix used on its name.

To reproduce the bug, just run this on the monitor:

 (qemu) host_net_add user name=foobar
 (qemu) host_net_remove 0 foobar
 invalid host network device foobar
 (qemu)

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6891 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFreeBSD host physical cdrom fixes
blueswir1 [Sat, 28 Mar 2009 08:37:13 +0000 (08:37 +0000)]
FreeBSD host physical cdrom fixes

This improves physical cdrom support on FreeBSD hosts to be almost as
good as on Linux, with the only notable exception that you still need to
either have the guest itself eject the disc if you want to take it
out/change it, or do a change command in the monitor after taking out
a disc in case a guest cannot eject it itself - otherwise the guest may
continue using state (like size) of the old disc.

Signed-off-by: Juergen Lock <nox@jelal.kn-bremen.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6888 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix warning in vl.c
blueswir1 [Sat, 28 Mar 2009 08:24:44 +0000 (08:24 +0000)]
Fix warning in vl.c

vl.c calls dma_helper_init, so it needs to include dma.h to get a
definition for it, otherwise we get compiler warnings like:

/home/hch/work/qemu/vl.c: In function 'main':
/home/hch/work/qemu/vl.c:5518: warning: implicit declaration of function 'dma_helper_init'

Signed-off-by: Christoph Hellwig <hch@lst.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6887 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix out of tree compilation
blueswir1 [Sat, 28 Mar 2009 08:14:42 +0000 (08:14 +0000)]
Fix out of tree compilation

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6886 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix compile problems with r6884
blueswir1 [Sat, 28 Mar 2009 08:13:56 +0000 (08:13 +0000)]
Fix compile problems with r6884

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6885 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSyncing documentation vs. -help vs. qemu_options table
blueswir1 [Sat, 28 Mar 2009 06:44:27 +0000 (06:44 +0000)]
Syncing documentation vs. -help vs. qemu_options table

Try to keep documentation about command line switches, -help text and
qemu_options table synchronized.

In true Qemu tradition, an include file is generated from single .hx file
containing all relevant information in one place. The include file is
parsed once for getting the enums, another time for getopt tables and
hird time for help messages. Texi documentation for the options is
generated from the same .hx file.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6884 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix absolute mouse events generated by SDL frontend.
balrog [Sat, 21 Mar 2009 01:09:16 +0000 (01:09 +0000)]
Fix absolute mouse events generated by SDL frontend.

In r6839 ("DisplayAllocator interface") the "width" and "height" globals
stopped ever being assigned.  Note that last time absolute input stopped
working was for the same reason.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6875 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoMove block dma helpers aiocb to store dma state (Avi Kivity)
aliguori [Fri, 20 Mar 2009 18:26:16 +0000 (18:26 +0000)]
Move block dma helpers aiocb to store dma state (Avi Kivity)

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

15 years agoUse vectored aiocb storage to store vector translation state (Avi Kivity)
aliguori [Fri, 20 Mar 2009 18:26:12 +0000 (18:26 +0000)]
Use vectored aiocb storage to store vector translation state (Avi Kivity)

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

15 years agoImplement cancellation method for dma async I/O (Avi Kivity)
aliguori [Fri, 20 Mar 2009 18:26:07 +0000 (18:26 +0000)]
Implement cancellation method for dma async I/O (Avi Kivity)

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

15 years agoConvert vectored aio emulation to use a dedicated pool (Avi Kivity)
aliguori [Fri, 20 Mar 2009 18:26:03 +0000 (18:26 +0000)]
Convert vectored aio emulation to use a dedicated pool (Avi Kivity)

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

15 years agoRefactor aio callback allocation to use an aiocb pool (Avi Kivity)
aliguori [Fri, 20 Mar 2009 18:25:59 +0000 (18:25 +0000)]
Refactor aio callback allocation to use an aiocb pool (Avi Kivity)

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

15 years agovirtio: Allow guest to defer VIRTIO_F_NOTIFY_ON_EMPTY (Alex Williamson)
aliguori [Fri, 20 Mar 2009 16:13:50 +0000 (16:13 +0000)]
virtio: Allow guest to defer VIRTIO_F_NOTIFY_ON_EMPTY (Alex Williamson)

There may be cases where the guest does not want the avail queue
interrupt, even when it's empty.  For the virtio-net case, the
guest may use a different buffering scheme or decide polling for
used buffers is more efficient.  This can be accomplished by simply
checking for whether the guest has acknowledged the existing notify
on empty flag.

Signed-off-by: Alex Williamson <alex.williamson@hp.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6865 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoe1000: Fix RX descriptor low threshold interrupt logic (Alex Williamson)
aliguori [Fri, 20 Mar 2009 16:13:47 +0000 (16:13 +0000)]
e1000: Fix RX descriptor low threshold interrupt logic (Alex Williamson)

The RXDMT0 interrupt is supposed to fire when the number of free
RX descriptors drops to some fraction of the total descriptors.
However in practice, it seems like we're adding this interrupt
cause on every RX.  Fix the logic to treat (tail - head) as the
number of free entries rather than the number of used entries.

Signed-off-by: Alex Williamson <alex.williamson@hp.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6864 c046a42c-6fe2-441c-8c8c-71466251a162