[PATCH] kvm-userspace: fix kvm initialization

A recent patch in qemu and its merge removed the availability of kvm_init_new_ap in target-$arch. The helper files now should call kvm_init_vcpu() directly. This patch changes that for powerpc. This should be true for target-ia64/op_helper.c too. Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTE

[PATCH] [mq]: fix-kvm-init.diff

diff --git a/qemu/target-ppc/helper.c b/qemu/target-ppc/helper.c --- a/qemu/target-ppc/helper.c +++ b/qemu/target-ppc/helper.c @@ -2959,10 +2959,8 @@ env->cpu_model_str = cpu_model; cpu_ppc_register_internal(env, def); cpu_ppc_reset(env); -#ifdef USE_KVM if (kvm_enabled()) -

[PATCH] [PATCH] kvm-userspace: fix io.h include merge confilict

From: Christian Ehrhardt <[EMAIL PROTECTED]> These two patches went in with a conflict: changeset: 16724:3ce8c651c948 user: Ehrhardt Christian <[EMAIL PROTECTED]> date:Tue Nov 04 12:10:51 2008 +0200 summary: kvm: qemu: move x86 specific calls introduced by device

[PATCH] kvm-userspace: add ibm archs to unifdef header

This adds powerpc to the unifdef header avi introduced. I think s390 should work the same way and is needed as they have libkvm already. Christian/Carsten could you ack that after a quick test if it is working? Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [diffstat] unifdef.h | 12

[PATCH] kvm-userspace: move x86 specific calls introduced by device assignement to x86 files

The device asignment patches added the x86 specific ioperm in qemu-kvm. This patch moves qemu-kvm.c:kvm_do_ioperm() to qemu-kvm-x86.c:kvm_arch_do_ioperm() The patch also changes the qemu-kvm header and the includes according to that. Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [

[PATCH] [PATCH] qemu: ppc: xer access prototypes no more used & implemented

From: Christian Ehrhardt <[EMAIL PROTECTED]> Revision 5500 of the qemu repository removed all code using ppc_load_xer & ppc_store_xer as well as their implementation. Another patch fixes it's usage in kvm-userspace for powerpc, but I think that header can now be cleaned up, therefore this patch t

[PATCH 05 of 10] [PATCH] libcflat: ppc: add timebase accessor

Provide a timebase accessor for ppc testcases. Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> [diffstat] config-powerpc-44x.mak |3 ++- test/lib/powerpc/44x/timebase.S | 28 test/lib/powerpc/44

[PATCH 04 of 10] [PATCH] user: ppc: implement PowerPC 44x libcflat

From: Hollis Blanchard <[EMAIL PROTECTED]> - Create a 44x-specific makefile. - Reorganize PowerPC makefiles to separate "simple" tests from those which link with libcflat. - Create a minimal libcflat testcase (which just exits). Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> Signed-off-by:

[PATCH 10 of 10] [PATCH] kvm-userspace: ppc: fix initial ppc memory setup

From: Christian Ehrhardt <[EMAIL PROTECTED]> The old memory initialization code was broken for all cases not fitting in one ram stick. This patch fixes the ram_stick calculation, now sets the proper base adresses per stick and removes the old workaround. Signed-off-by: Christian Ehrhardt <[EMAIL

[PATCH 08 of 10] [PATCH] qemu: ppc: if not a uImage, try to load kernel as ELF

From: Hollis Blanchard <[EMAIL PROTECTED]> This allows qemu to load "bare metal" ELF kernels, useful for standalone benchmarks and testcases. We could/should also load the specified file as a flat binary, if both uImage and ELF loaders fail. (See hw/arm_boot.c.) Signed-off-by: Hollis Blanchard <

[PATCH 02 of 10] [PATCH] user: ppc: fix threading bugs in main-ppc.c

From: Hollis Blanchard <[EMAIL PROTECTED]> - call io_table_register() before any vcpus have started - wait for all vcpus to exit before exiting the parent thread Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [diffstat] main-ppc.c |

[PATCH 07 of 10] [PATCH] qemu: ppc: define maximum SMP limit as 1 for Bamboo

From: Christian Ehrhardt <[EMAIL PROTECTED]> Fix for qemu runtime error. Full error message: Number of SMP cpus requested (1), exceeds max cpus supported by machine `bamboo' (0) Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> --- [diffst

[PATCH 06 of 10] [PATCH] user: ppc: add stub nmi handler

From: Hollis Blanchard <[EMAIL PROTECTED]> Adding a nmi stub handler for user/main-ppc.c Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [diffstat] main-ppc.c |6 ++ 1 file changed, 6 insertions(+) [diff] diff --git a/user/

[PATCH 03 of 10] [PATCH] user: ppc: better error reporting in load_file

From: Hollis Blanchard <[EMAIL PROTECTED]> Fancy description. Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [diffstat] main-ppc.c |3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) [diff] diff --git a/user/main-ppc.c b/us

[PATCH 09 of 10] [PATCH] kvm: external module: Treat NONARCH_CONFIG as a list

From: Hollis Blanchard <[EMAIL PROTECTED] As discussed on the list the unifdef changes break powerpc (and more ?). A fix is to treat NONARCH_CONFIG as a list instead of a single item. Signed-off-by: Hollis Blanchard <[EMAIL PROTECTED]> Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [d

[PATCH 00 of 10] kvm-userspace: ppc: userspace fixes for powerpc

From: Christian Ehrhardt <[EMAIL PROTECTED]> This is a set of various fixes in kvm-userspace for powerpc. This time without the split between user/* and the rest and without the qemu patch (sent separate to qemu-devel now). Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> b/kernel/Makefile

[PATCH 01 of 10] [PATCH] kvm-userspace: powerpc: fix env->xer access

From: Christian Ehrhardt <[EMAIL PROTECTED]> Since qemu revision 5500 which was merged with the last qemu merge env->xer is accessed directly. Signed-off-by: Christian Ehrhardt <[EMAIL PROTECTED]> --- [diffstat] qemu-kvm-powerpc.c |4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) [d