qemu
15 years agoMerge commit 'origin/master' into maemo-test
Riku Voipio [Thu, 5 Mar 2009 09:14:29 +0000 (11:14 +0200)]
Merge commit 'origin/master' into maemo-test

15 years agoRevert "flush stdout after printing usage()"
Riku Voipio [Thu, 5 Mar 2009 09:14:13 +0000 (11:14 +0200)]
Revert "flush stdout after printing usage()"

This reverts commit 7f0ceaa78b572f3fab0627c3a084ce1173909e85.

15 years agoAdd version information for 0.10.0 release.
aliguori [Wed, 4 Mar 2009 22:47:59 +0000 (22:47 +0000)]
Add version information for 0.10.0 release.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6685 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix "info registers" under kvm.
balrog [Wed, 4 Mar 2009 21:00:07 +0000 (21:00 +0000)]
Fix "info registers" under kvm.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6684 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoUpdate cocoa.m to match new DisplayState code (Samuel Benson)
aliguori [Wed, 4 Mar 2009 19:25:22 +0000 (19:25 +0000)]
Update cocoa.m to match new DisplayState code (Samuel Benson)

Version 2 does as follows:

[1]: Corrects endianness on issues by using native BGR to RGB conversion
[2]: Uses DisplayState accessors for obtaining graphics context information,
     which
[3]: Removes now unused variables, and
[4]: Allows reading of varying color modes (32bit/24/16), and converting to
     native colorspace
[5]: Attempts to keep itself centered on screen (as opposed to bottom right,
     which immediately goes off screen after bios load) on context changes
    (window resizes)

Testing working on i386 (gentoo, Windows 2000) and PPC (debian) guests on PPC
and x86 Macs.

In regards to [4], Windows 2000 displays fine on quick tests, but on the lowest
setting I could test, 16bit color depth at 4bpp, colors are slightly off. I
used gentoo install-x86-minimal-2008.0 in framebuffer mode to test above
setting; the usual grey text is now blue, and Tux appears to be BGR shifted. I
do not know if previous code worked at such a low color setting.

Signed-off-by: Samuel Benson <qemu_ml@digitalescape.info>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6683 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoThe different VFP variants will never occur at the same time, and giving them
pbrook [Wed, 4 Mar 2009 12:18:15 +0000 (12:18 +0000)]
The different VFP variants will never occur at the same time, and giving them
different names confuses GDB, so use org.gnu.gdb.arm.vfp for all of them.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6682 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDB-DMA IDE asynchronous I/O
aurel32 [Wed, 4 Mar 2009 07:20:40 +0000 (07:20 +0000)]
DB-DMA IDE asynchronous I/O

Signed-off-by: Laurent Vivier <Laurent@vivier.eu>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6681 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix slavio_intctl.c compile errors (Robert Reif)
blueswir1 [Tue, 3 Mar 2009 20:11:43 +0000 (20:11 +0000)]
Fix slavio_intctl.c compile errors (Robert Reif)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6680 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoCRIS: Avoid endless loops for unstested CRIS cpu-state load/save.
edgar_igl [Tue, 3 Mar 2009 18:07:11 +0000 (18:07 +0000)]
CRIS: Avoid endless loops for unstested CRIS cpu-state load/save.

Fix several endless loops.

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Acked-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6679 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix SDL on evdev hosts (Anthony Liguori)
aliguori [Tue, 3 Mar 2009 17:37:21 +0000 (17:37 +0000)]
Fix SDL on evdev hosts (Anthony Liguori)

This patch corrects SDL support on X11 hosts using evdev.  It's losely based
on the previous patch by Dustin Kirkland and the evdev support code in gtk-vnc
written by Daniel Berrange.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6678 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix CVE-2008-0928 - insufficient block device address range checking (Anthony Liguori)
aliguori [Tue, 3 Mar 2009 17:37:16 +0000 (17:37 +0000)]
Fix CVE-2008-0928 - insufficient block device address range checking (Anthony Liguori)

Introduce a growable flag that's set by bdrv_file_open().  Block devices should
never be growable, only files that are being used by block devices.

I went through Fabrice's early comments about the patch that was first applied.
While I disagree with that patch, I also disagree with Fabrice's suggestion.

There's no good reason to do the checks in the block drivers themselves.  It
just increases the possibility that this bug could show up again.  Since we're
calling bdrv_getlength() to determine the length, we're giving the block drivers
a chance to chime in and let us know what range is valid.

Basically, this patch makes the BlockDriver API guarantee that all requests are
within 0..bdrv_getlength() which to me seems like a Good Thing.

What do others think?

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6677 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDB-DMA cleanup
aurel32 [Tue, 3 Mar 2009 09:14:10 +0000 (09:14 +0000)]
DB-DMA cleanup

Signed-off-by: Laurent Vivier <Laurent@vivier.eu>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6676 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH4: Fixed last UTLB unused and URB/URC management
aurel32 [Tue, 3 Mar 2009 09:14:01 +0000 (09:14 +0000)]
SH4: Fixed last UTLB unused and URB/URC management

Signed-off-by: Lionel Landwerlin <lionel.landwerlin@openwide.fr>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6675 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agofix SIGSEGV when USE_KQEMU is defined
aurel32 [Tue, 3 Mar 2009 06:28:26 +0000 (06:28 +0000)]
fix SIGSEGV when USE_KQEMU is defined

Signed-off-by: Alex Ivanov <void@aleksoft.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6674 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agosh4: r2d fix no ide/net case
aurel32 [Tue, 3 Mar 2009 06:23:17 +0000 (06:23 +0000)]
sh4: r2d fix no ide/net case

Fix invalid access/crash when there is no IDE device or NET device.

Signed-off-by: Takashi YOSHII <takasi-y@ops.dti.ne.jp>
Tested-by: Shin-ichiro KAWASAKI <kawasaki@juno.dti.ne.jp>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6673 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH4: Fixed last UTLB unused
aurel32 [Tue, 3 Mar 2009 06:14:43 +0000 (06:14 +0000)]
SH4: Fixed last UTLB unused

Version 2 of the patch.

Signed-off-by: Lionel Landwerlin <lionel.landwerlin@openwide.fr>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6672 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH4: Fixed last UTLB unused
aurel32 [Tue, 3 Mar 2009 06:12:34 +0000 (06:12 +0000)]
SH4: Fixed last UTLB unused

With my previous patch (the one monitoring tlb), I found that the last
TLB entry was never use. Here a little fix.

Signed-off-by: Lionel Landwerlin <lionel.landwerlin@openwide.fr>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6671 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH4: Added monitoring of TLBs
aurel32 [Tue, 3 Mar 2009 06:12:22 +0000 (06:12 +0000)]
SH4: Added monitoring of TLBs

This patch might interest some people trying (as I try to do) to fix
some tlbs for kernel/user space data sharing.

Signed-off-by: Lionel Landwerlin <lionel.landwerlin@openwide.fr>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6670 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agotarget-ppc: improve mfcr/mtcrf
aurel32 [Tue, 3 Mar 2009 06:12:14 +0000 (06:12 +0000)]
target-ppc: improve mfcr/mtcrf

- use ctz32 instead of ffs - 1
- small optimisation of mtcrf
- add the name of both opcodes

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6669 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoclean build: Fix remaining sh4 warnings
aurel32 [Tue, 3 Mar 2009 06:12:03 +0000 (06:12 +0000)]
clean build: Fix remaining sh4 warnings

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@6668 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix mtcrf/mfcr
malc [Mon, 2 Mar 2009 22:39:39 +0000 (22:39 +0000)]
Fix mtcrf/mfcr

Noticed by Alexander Graf

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6667 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSH: Implement MOVCO.L and MOVLI.L
aurel32 [Mon, 2 Mar 2009 17:13:21 +0000 (17:13 +0000)]
SH: Implement MOVCO.L and MOVLI.L

* target-sh4/cpu.h (struct CPUSH4State): New field ldst.
* target-sh4/translate.c (cpu_ldst): New.
  (sh4_translate_init): Initialize cpu_ldst.
  (_decode_opc): Support MOVCO.L and MOVLI.L.

(Vladimir Prus)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6666 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSet mime-type of pc-bios/mpc8544ds.dtb to application/octet-stream
aurel32 [Mon, 2 Mar 2009 17:10:47 +0000 (17:10 +0000)]
Set mime-type of pc-bios/mpc8544ds.dtb to application/octet-stream

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6665 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agokvm/powerpc: flat device tree files for MPC8544DS
aurel32 [Mon, 2 Mar 2009 16:42:49 +0000 (16:42 +0000)]
kvm/powerpc: flat device tree files for MPC8544DS

mpc8544ds.dts is copied from kerenl.
Non-supported devices has been removed.

Signed-off-by: Liu Yu <yu.liu@freescale.com>
Acked-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6664 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agokvm/powerpc: Add MPC8544DS board support
aurel32 [Mon, 2 Mar 2009 16:42:42 +0000 (16:42 +0000)]
kvm/powerpc: Add MPC8544DS board support

This patch add an emulation of MPC8544DS board.
It can work on All E500 platforms.

Signed-off-by: Liu Yu <yu.liu@freescale.com>
Acked-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6663 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agokvm/powerpc: Add irq support for E500 core
aurel32 [Mon, 2 Mar 2009 16:42:32 +0000 (16:42 +0000)]
kvm/powerpc: Add irq support for E500 core

Signed-off-by: Liu Yu <yu.liu@freescale.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6662 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agokvm/powerpc: Add freescale pci controller's support
aurel32 [Mon, 2 Mar 2009 16:42:23 +0000 (16:42 +0000)]
kvm/powerpc: Add freescale pci controller's support

This patch add the emulation of freescale's pci controller for MPC85xx platform.

Signed-off-by: Liu Yu <yu.liu@freescale.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6661 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agohw/openpic: define list in mpic_init() const
aurel32 [Mon, 2 Mar 2009 16:42:14 +0000 (16:42 +0000)]
hw/openpic: define list in mpic_init() const

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6660 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agokvm/powerpc: Enable MPIC for E500 platform.
aurel32 [Mon, 2 Mar 2009 16:42:04 +0000 (16:42 +0000)]
kvm/powerpc: Enable MPIC for E500 platform.

MPIC and OpenPIC have very similar design.
So a lot of code can be reused.

Signed-off-by: Liu Yu <yu.liu@freescale.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6659 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoUpdate to OpenBIOS 1.0
blueswir1 [Sun, 1 Mar 2009 20:18:09 +0000 (20:18 +0000)]
Update to OpenBIOS 1.0

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6658 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFlush stdout after printing usage()
blueswir1 [Sat, 28 Feb 2009 20:14:00 +0000 (20:14 +0000)]
Flush stdout after printing usage()

Fixes qemu-arm|grep cpu - with _exit() getting output from qemu --help
is a bit random. Since no atexit() handlers are registered for user mode
emulation, just use exit() instead.

Signed-off-by: Riku Voipio <riku.voipio@iki.fi>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6657 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoImplement HIOR
blueswir1 [Sat, 28 Feb 2009 18:39:42 +0000 (18:39 +0000)]
Implement HIOR

A real 970 CPU starts up with HIOR=0xfff00000 and triggers a reset
exception, basically ending up at IP 0xfff001000.

Later on this HIOR has to be set to 0 by the firmware in order to
enable the OS to handle interrupts on its own.

This patch maps HIOR to exec_prefix, which does the same thing
internally in qemu already.

It replaces the previous patch that changed the 970 initialization
constants, as this is the clean solution to the same problem.

Signed-off-by: Alexander Graf <alex@csgraf.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6656 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoChange default werror semantics from "report" to "enospc"
aliguori [Sat, 28 Feb 2009 16:51:01 +0000 (16:51 +0000)]
Change default werror semantics from "report" to "enospc"

Practically speaking, "report" causes a lot of issues when encountering a host
ENOSPC error.  Switch to "enospc" as the default werror semantics.  All host
errors other than ENOSPC will be reported to the guest.  ENOSPC will cause the
VM to stop.  Asynchronous notifications are needed to inform management tools
that some action should be taken but stopping the VM is at least better than
undefined behavior in the guest.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6655 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix typo in gen_qemu_ld32s
blueswir1 [Sat, 28 Feb 2009 08:25:29 +0000 (08:25 +0000)]
Fix typo in gen_qemu_ld32s

When the CPU is in little endian mode, it should load values from RAM
in byte swapped manner. This check is in all the ld and st functions,
but misspelled in gen_qemu_ld32s.

This patch fixes the misspelling and makes ppc64 Linux happier.

Signed-off-by: Alexander Graf <alex@csgraf.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6654 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoRemove some warnings and fix windows build.
aliguori [Fri, 27 Feb 2009 22:16:47 +0000 (22:16 +0000)]
Remove some warnings and fix windows build.

Initialize some variables to make GCC happy and switch from using index to
strchr.  index is not available on Windows.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6653 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agochroot and change user support (Nolan)
aliguori [Fri, 27 Feb 2009 22:09:45 +0000 (22:09 +0000)]
chroot and change user support (Nolan)

Resent with fixed formatting.

This patch adds two new command line options:
 -chroot <dir>
 -runas <user>

This is useful for running qemu as an unprivileged user in a chroot
jail.  To avoid having to populate the jail, chrooting happens right
before the start of guest execution.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6652 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoUnify default parallel console size (Jan Kiszka)
aliguori [Fri, 27 Feb 2009 20:14:29 +0000 (20:14 +0000)]
Unify default parallel console size (Jan Kiszka)

Serial and virtio consoles already use 80x24 characters as default size,
apply the same to the parallel port consoles.

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@6651 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAllow additions of ACPI tables from command line (Gleb Natapov)
aliguori [Fri, 27 Feb 2009 20:12:36 +0000 (20:12 +0000)]
Allow additions of ACPI tables from command line (Gleb Natapov)

This is needed to dynamically add SLIC tables with Windows
activation keys.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6650 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agox86: use qemu_log_mask on triple faults (Chris Wright)
aliguori [Fri, 27 Feb 2009 20:05:13 +0000 (20:05 +0000)]
x86: use qemu_log_mask on triple faults (Chris Wright)

replace open coded qemu_log_mask with proper macro

Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6649 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDo not enable a default virtio console
aliguori [Fri, 27 Feb 2009 20:01:39 +0000 (20:01 +0000)]
Do not enable a default virtio console

This upsets Windows installs and right now, virtio console isn't very useful
as a default device.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6648 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agonet socket verify packet size (Dustin Kirkland)
aliguori [Fri, 27 Feb 2009 19:54:01 +0000 (19:54 +0000)]
net socket verify packet size (Dustin Kirkland)

net socket oversized packet

This is a patch being carried by Ubuntu against kvm/qemu.

Verify packet size before performing memcpy().

Signed-off-by: Dustin Kirkland <kirkland@canonical.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6647 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoMissing cirrus_invalidate_region() in cirrus_do_copy() (Brian Kress)
aliguori [Fri, 27 Feb 2009 19:53:57 +0000 (19:53 +0000)]
Missing cirrus_invalidate_region() in cirrus_do_copy() (Brian Kress)

After doing a video to video copy, cirrus_do_copy() in cirrus_vga.c does a
conditional call to cirrus_invalidate_region() with an "if (!notify)" test.
However at this point the blt has already been done so it seems like this call
should be made unconditionally.  The patch below fixes some display update
problems when dragging windows under both X (fedora 10 guest) and a Windows XP
guest.

Signed-off-by: Brian Kress <kressb@moose.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6646 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agofix pci net hot-remove (Marcelo Tosatti)
aliguori [Thu, 26 Feb 2009 16:40:31 +0000 (16:40 +0000)]
fix pci net hot-remove (Marcelo Tosatti)

Missing brackets, doh.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6645 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agofix raw_aio_remove (Stefano Stabellini)
aliguori [Thu, 26 Feb 2009 16:40:19 +0000 (16:40 +0000)]
fix raw_aio_remove (Stefano Stabellini)

Hi all,
this small patch fixes a bug in the list iteration of raw_aio_remove.
Cheers,

Stefano

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6644 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix ColdFire fmovem. Free the temporary we just allocated rather than some
pbrook [Tue, 24 Feb 2009 22:17:35 +0000 (22:17 +0000)]
Fix ColdFire fmovem.  Free the temporary we just allocated rather than some
other arbitrary value (e.g. A7).

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6643 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoReally fix ColdFire fdmoved address decoding.
pbrook [Mon, 23 Feb 2009 15:19:34 +0000 (15:19 +0000)]
Really fix ColdFire fdmoved address decoding.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6642 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix ColdFire fdmoved address decoding.
pbrook [Mon, 23 Feb 2009 14:17:33 +0000 (14:17 +0000)]
Fix ColdFire fdmoved address decoding.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6641 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoDarwin: Check for x86_64 only on i386
malc [Mon, 23 Feb 2009 14:11:10 +0000 (14:11 +0000)]
Darwin: Check for x86_64 only on i386

The sysctl variable if we're 64-bit capable only exists on i386. So we should only check it if we're on i386.

This suppresses a warning on PowerPC spotted by Andreas Faerber.

Signed-off-by: Alexander Graf <alex@csgraf.de>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6640 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix unassigned region offsets.
pbrook [Mon, 23 Feb 2009 13:16:07 +0000 (13:16 +0000)]
Fix unassigned region offsets.

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6639 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoCRIS: Fix remaining build warnings.
edgar_igl [Sun, 22 Feb 2009 11:59:59 +0000 (11:59 +0000)]
CRIS: Fix remaining build warnings.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Acked-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6638 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoTurn MMU off on reset
blueswir1 [Sat, 21 Feb 2009 17:29:14 +0000 (17:29 +0000)]
Turn MMU off on reset

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6637 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoTurn MMUs and caches off on reset
blueswir1 [Sat, 21 Feb 2009 11:13:51 +0000 (11:13 +0000)]
Turn MMUs and caches off on reset

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6636 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoReload ptimers on reset
blueswir1 [Sat, 21 Feb 2009 11:13:10 +0000 (11:13 +0000)]
Reload ptimers on reset

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6635 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoProperly handle pthread_cond_timedwait timing out
malc [Sat, 21 Feb 2009 05:48:19 +0000 (05:48 +0000)]
Properly handle pthread_cond_timedwait timing out

pthread_cond_timedwait is allowed to both consume the signal and
return with the value indicating the timeout, hence predicate should
always be (re)checked before taking an action

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6634 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoCosmetics
malc [Sat, 21 Feb 2009 05:48:17 +0000 (05:48 +0000)]
Cosmetics

Avoid repeated creation/initalization/destruction of attr and calls to
getpid

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6633 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAvoid thundering herd problem
malc [Sat, 21 Feb 2009 05:48:15 +0000 (05:48 +0000)]
Avoid thundering herd problem

Broadcast was used so that the I/O threads would wakeup, reset their
ts values and all but one go to sleep, in other words an optimization
to prevent threads from exiting in presence of continuing I/O
activity. Spurious wakeups make the looping around cond_timedwait with
ever reinitialized ts potentially unsafe and as such ts in no longer
reinitilized inside the loop, hence switch to signal is warranted and
this benefits of this particlaur optimization are lost.

(It's worth noting that timed variants of pthread calls use realtime
clock by default, and therefore can hang "forever" should the host
time be changed. Unfortunatelly not all host systems QEMU runs on
support CLOCK_MONOTONIC and/or pthread_condattr_setclock with this
value)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6632 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAvoid infinite loop around timed condition variable
malc [Sat, 21 Feb 2009 05:48:13 +0000 (05:48 +0000)]
Avoid infinite loop around timed condition variable

This can happen due to spurious wakeups

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6631 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoError checking
malc [Sat, 21 Feb 2009 05:48:11 +0000 (05:48 +0000)]
Error checking

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6630 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoport omap3 code to post-displaystate
Riku Voipio [Fri, 20 Feb 2009 10:52:24 +0000 (12:52 +0200)]
port omap3 code to post-displaystate

15 years agoFix branch debugging
blueswir1 [Thu, 19 Feb 2009 20:17:09 +0000 (20:17 +0000)]
Fix branch debugging

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6629 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agofix merge conflict
Riku Voipio [Thu, 19 Feb 2009 16:20:58 +0000 (18:20 +0200)]
fix merge conflict

15 years agomake omap_dss work on both omap2 and omap3
Riku Voipio [Thu, 19 Feb 2009 16:17:50 +0000 (18:17 +0200)]
make omap_dss work on both omap2 and omap3

15 years agoRevert "Add BGR format host frame buffer support in Epson LCD controller emulation"
Riku Voipio [Thu, 19 Feb 2009 15:32:21 +0000 (17:32 +0200)]
Revert "Add BGR format host frame buffer support in Epson LCD controller emulation"

This reverts commit 7384e7f8cd34dac587e947fe94745aacfda4dd0d.

15 years agorevert multi-omap for omap1/omap2
Riku Voipio [Thu, 19 Feb 2009 15:03:56 +0000 (17:03 +0200)]
revert multi-omap for omap1/omap2

15 years agorevert multi-omap for omap1/omap2
Riku Voipio [Thu, 19 Feb 2009 15:03:56 +0000 (17:03 +0200)]
revert multi-omap for omap1/omap2

15 years agoMerge branch 'upstream' into maemo-test
Riku Voipio [Thu, 19 Feb 2009 14:52:18 +0000 (16:52 +0200)]
Merge branch 'upstream' into maemo-test

Resolve conflicts

Conflicts:

.gitignore
hw/omap_dss.c
linux-user/main.c
linux-user/syscall.c

15 years agoMerge branch 'juha-devel'
Riku Voipio [Thu, 19 Feb 2009 11:58:35 +0000 (13:58 +0200)]
Merge branch 'juha-devel'

15 years agoFix cpu_physical_memory_rw() for 64-bit I/O accesses
aurel32 [Wed, 18 Feb 2009 21:37:17 +0000 (21:37 +0000)]
Fix cpu_physical_memory_rw() for 64-bit I/O accesses

KVM uses cpu_physical_memory_rw() to access the I/O devices. When a
read or write with a length of 8-byte is requested, it is split into 2
4-byte accesses.

This has been broken in revision 5849. After this revision, only the
first 4 bytes are actually read/write to the device, as the target
address is changed, so on the next iteration of the loop the next 4
bytes are actually read/written elsewhere (in the RAM for the graphic
card).

This patch fixes screen corruption (and most probably data corruption)
with FreeBSD/amd64. Bug #2556746 in KVM bugzilla.

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

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6628 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoAvoid running audio ctl's when vm is not running
malc [Wed, 18 Feb 2009 20:44:04 +0000 (20:44 +0000)]
Avoid running audio ctl's when vm is not running

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6627 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoclear ERRI bit in MMCHS_STAT so it accurately reflects the rest error bits
Riku Voipio [Wed, 18 Feb 2009 15:30:40 +0000 (17:30 +0200)]
clear ERRI bit in MMCHS_STAT so it accurately reflects the rest error bits

15 years agoDummy support for CMD_TYPE bits in OMAP3 MMC CMD register
Juha Riihimäki [Wed, 18 Feb 2009 13:49:42 +0000 (15:49 +0200)]
Dummy support for CMD_TYPE bits in OMAP3 MMC CMD register

This allows the Linux kernel to boot a bit further as it is issuing CMD12 with "I/O Abort" CMD_TYPE. However as this is not correctly emulated yet it seems the kernel gets trash instead of the real stuff...

15 years agoFix for OMAP3 MMC FIFO handling
Juha Riihimäki [Wed, 18 Feb 2009 13:03:11 +0000 (15:03 +0200)]
Fix for OMAP3 MMC FIFO handling

DMA transfers are now stopped correctly.

15 years agoFix DMA channel numbers for OMAP3
Juha Riihimäki [Wed, 18 Feb 2009 12:22:22 +0000 (14:22 +0200)]
Fix DMA channel numbers for OMAP3

All DMA channel numbers were off by one... :)

15 years agoOMAP DMA debug printout fix, part2
Juha Riihimäki [Wed, 18 Feb 2009 10:32:16 +0000 (12:32 +0200)]
OMAP DMA debug printout fix, part2

Fix per-channel register number printouts also for omap_dma4_read function.

15 years agoFix OMAP DMA4 debug printouts
Juha Riihimäki [Wed, 18 Feb 2009 10:24:03 +0000 (12:24 +0200)]
Fix OMAP DMA4 debug printouts

Illegal access in per-channel register region prints out the register numbers relative to channel number, which is not printed out. Changed printing to display register numbers relative to the DMA controller region.

15 years agoFix block count for OMAP3 MMC emulation
Juha Riihimäki [Wed, 18 Feb 2009 09:28:59 +0000 (11:28 +0200)]
Fix block count for OMAP3 MMC emulation

Block count was extracted wrongly from a write to MMCHS_BLK register. Also rearranged the code a bit to resemble more the similar implementation done for OMAP2.

15 years agoDisable extensive SD debug printouts
Juha Riihimäki [Wed, 18 Feb 2009 06:30:33 +0000 (08:30 +0200)]
Disable extensive SD debug printouts

Enough is enough and too much is too much

15 years agoMerge commit 'origin/riku-devel' into juha-devel
Juha Riihimäki [Wed, 18 Feb 2009 06:06:35 +0000 (08:06 +0200)]
Merge commit 'origin/riku-devel' into juha-devel

15 years agoEnable debug traces in SD emulation
Juha Riihimäki [Wed, 18 Feb 2009 06:05:11 +0000 (08:05 +0200)]
Enable debug traces in SD emulation

Added some more debug information in the SD emulation and fixed OMAP3 MMC2/3 initialization with correct NULL values.

15 years agodo not pretend to support low voltage operation
Riku Voipio [Tue, 17 Feb 2009 16:16:26 +0000 (18:16 +0200)]
do not pretend to support low voltage operation

eliminate "mmc0: SD card claims to support the incompletely defined 'low voltage
range'. This will be ignored." warning

15 years agoYet-Another-Fix for OMAP3 I2C interrupt emulation
Juha Riihimäki [Tue, 17 Feb 2009 12:35:36 +0000 (14:35 +0200)]
Yet-Another-Fix for OMAP3 I2C interrupt emulation

It Is Now Fixed(tm).

15 years agoFix XRDY/RRDY/XDR/RDR interrupt generation for OMAP3 I2C
Juha Riihimäki [Tue, 17 Feb 2009 11:15:54 +0000 (13:15 +0200)]
Fix XRDY/RRDY/XDR/RDR interrupt generation for OMAP3 I2C

Patch for XRDY/RRDY/XDR/RDR interrupt generation in the OMAP3 I2C emulation.

15 years agoOMAP3 HS MMC emulation interrupt fix and other small stuff
Juha Riihimäki [Tue, 17 Feb 2009 08:42:25 +0000 (10:42 +0200)]
OMAP3 HS MMC emulation interrupt fix and other small stuff

The OMAP3 HS MMC controller interrupt state is now correctly updated after a write to STAT register. Hardcoded OMAP SDRC to always return "no address error" to an error query. Changed CR+LF line terminators to LF in a number of files.

15 years agomore gpio clocks and rtc stub
Riku Voipio [Mon, 16 Feb 2009 15:59:33 +0000 (17:59 +0200)]
more gpio clocks and rtc stub

15 years agovirtio: Remove malloc failure checks (Jan Kiszka)
aliguori [Mon, 16 Feb 2009 15:47:27 +0000 (15:47 +0000)]
virtio: Remove malloc failure checks (Jan Kiszka)

No need to check for failing qemu_malloc anymore.

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@6626 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoGenerate PCI hotplug interrupt only if corespondent EN bit is set. (Gleb Natapov)
aliguori [Mon, 16 Feb 2009 15:36:08 +0000 (15:36 +0000)]
Generate PCI hotplug interrupt only if corespondent EN bit is set. (Gleb Natapov)

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6625 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoFix GPE registers read/write handling. (Gleb Natapov)
aliguori [Mon, 16 Feb 2009 15:36:03 +0000 (15:36 +0000)]
Fix GPE registers read/write handling. (Gleb Natapov)

For STS register bit are cleared by writing 1 into it.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6624 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agospecify vmchannel as a net option (Gleb Natapov)
aliguori [Mon, 16 Feb 2009 15:34:18 +0000 (15:34 +0000)]
specify vmchannel as a net option (Gleb Natapov)

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6623 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoMerge commit 'juri/juha-devel' into test
Riku Voipio [Mon, 16 Feb 2009 15:01:35 +0000 (17:01 +0200)]
Merge commit 'juri/juha-devel' into test

15 years agomore twl regs
Riku Voipio [Mon, 16 Feb 2009 15:01:26 +0000 (17:01 +0200)]
more twl regs

15 years agoFix hardware accelerated video to video copy on Cirrus VGA (Brian Kress)
aliguori [Mon, 16 Feb 2009 14:59:35 +0000 (14:59 +0000)]
Fix hardware accelerated video to video copy on Cirrus VGA (Brian Kress)

cirrus_do_copy() in hw/cirrus_vga.c seems to make some incorrect assumptions
about video memory layout.  It tries to convert addresses to coordinates
assuming that one row of data is (width * depth) bytes long.  The correct way
seems to be to use the pitch fields in the CirrusVGAState structure instead.

Without this patch, I get lots of screen corruption when I try to drag a window
under X as it's passing the wrong coordinates to the display surface for the
copy.  With this patch I can drag a window with no screen corruption.

Signed-off-by: Brian Kress <kressb@moose.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6622 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoSupport multiple VNC clients (Brian Kress)
aliguori [Mon, 16 Feb 2009 14:59:30 +0000 (14:59 +0000)]
Support multiple VNC clients (Brian Kress)

Change structure associated with a display from VncState to a new structure
VncDisplay. Remove client specific fields from VncDisplay. Remove display
specific fields from VncState. Maintain a linked list of VncStates per
VncDisplay structure, update as necessary. When updates/resizes/copies come in
from the hardware, dispatch to all clients.

Signed-off-by: Brian Kress <kressb@moose.net>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6621 c046a42c-6fe2-441c-8c8c-71466251a162

15 years agoOMAP3 I2C patch for FIFO draining feature handling
Juha Riihimäki [Mon, 16 Feb 2009 14:44:03 +0000 (16:44 +0200)]
OMAP3 I2C patch for FIFO draining feature handling

XDR/RDR STAT bits are set only when corresponding IE bits are set. Therefore polling for XDR/RDR bits is now impossible. However this seems to make the Linux kernel feel better.

15 years agoMerge commit 'origin/riku-devel' into juha-devel
Juha Riihimäki [Mon, 16 Feb 2009 13:25:52 +0000 (15:25 +0200)]
Merge commit 'origin/riku-devel' into juha-devel

15 years agoMerge commit 'juri/riku-devel' into test
Riku Voipio [Mon, 16 Feb 2009 13:20:11 +0000 (15:20 +0200)]
Merge commit 'juri/riku-devel' into test

15 years agoMoved OMAP3 boot ROM MMC loader emulation from beagle.c to omap3.c
Juha Riihimäki [Mon, 16 Feb 2009 13:19:47 +0000 (15:19 +0200)]
Moved OMAP3 boot ROM MMC loader emulation from beagle.c to omap3.c

Since the MMC loader emulation of OMAP3 boot ROM is generic for OMAP3 and not specific for Beagleboard, moved the relevant functions to omap3.c from beagle.c.

15 years agomulti-mmc support
Riku Voipio [Mon, 16 Feb 2009 13:02:34 +0000 (15:02 +0200)]
multi-mmc support

15 years agomulti-mmc support
Riku Voipio [Mon, 16 Feb 2009 13:02:34 +0000 (15:02 +0200)]
multi-mmc support

15 years agoSome improvements for OMAP3
Juha Riihimäki [Mon, 16 Feb 2009 12:08:38 +0000 (14:08 +0200)]
Some improvements for OMAP3

Add support for PRCM IRQ, SDRC DLLA_STATUS register. Mirrors modifications to CM ICLKEN1 and ICLKEN3 to corresponding idle status registers.

15 years agoReplace numeric constants with symbols for OMAP3 L4
Juha Riihimäki [Mon, 16 Feb 2009 09:13:13 +0000 (11:13 +0200)]
Replace numeric constants with symbols for OMAP3 L4

OMAP3 L4 region ids and target agents now have symbolic names instead of plain numbers. This makes it much easier to a) read the code and b) rearrange the order of the data structures due to e.g. inserting new known regions or agents.