kvm-devel — kernel virtual machine development

Re: [kvm-devel] More queries on guest swapping in KVM

[Izik E. <iz...@qu...>]

Arn wrote:
> Hi ,
>
> Regarding the guest swapping mechanism (host ability to swap out guest 
> pages),
>
> 1) Can the host swap out any page of a guest VM ? (including the 
> guest's kernel pages)

right now kvm can swap any pages but pages that are mapped in it mmu 
cache. (it doesnt matther if it kernel page or not)]
with mmu notifiers witch probably be marged to 2.6.26 we will be able to 
swap the entire pages of the guest.

>
> 2) Does the guest know its page has been swapped out or is this 
> transparent to the guest ? ( I think the guest
> wouldn't know but I just want to be sure)

it doesnt know.

>
> 3) Is there any way the guest can prevent the host from swapping out 
> its pages ?

not without writing new code.

>
> thanks
> Arn
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save $100. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> ------------------------------------------------------------------------
>
> _______________________________________________
> kvm-devel mailing list
> kvm...@li...
> https://lists.sourceforge.net/lists/listinfo/kvm-devel
>   

[kvm-devel] [ kvm-Bugs-1881418 ] win xp guest BSOD and host kernel oops if "-smb path"

[SourceForge.net <no...@so...>]

Bugs item #1881418, was opened at 2008-01-28 21:55
Message generated for change (Comment added) made by fishor-flex
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1881418&group_id=180599

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: intel
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Alexey Fisher (fishor-flex)
Assigned to: Nobody/Anonymous (nobody)
Summary: win xp guest BSOD and host kernel oops if "-smb path"

Initial Comment:
Windows XP SP2 guest will get blueskreen and kernel ( Linux zwerg 2.6.24-rc6-g933400f4 #3 SMP PREEMPT ) will oops somtime with complete system freez  if smb share cenabled. There is no difference if "-smb /home/path" or "-smb share" used.

kvm-kernel commit version b90a99cf8e03e96a4e3347d9e9ee3750e53d07fd

kvm-userspace e82b1fc2be848acc537512cf8e00b9aea37f3fb2



----------------------------------------------------------------------

>Comment By: Alexey Fisher (fishor-flex)
Date: 2008-05-01 07:50

Message:
Logged In: YES 
user_id=1409122
Originator: YES

The dmesg was attached for a long time. See bottom of this page

https://sourceforge.net/tracker/download.php?group_id=180599&atid=893831&file_id=264057&aid=1881418

----------------------------------------------------------------------

Comment By: Marcelo Tosatti (mtosatti)
Date: 2008-04-30 18:56

Message:
Logged In: YES 
user_id=2022487
Originator: NO

Can you please provide the host kernel oops? 

----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-29 18:29

Message:
Logged In: YES 
user_id=1409122
Originator: YES

with last git update ( 29.01.08 18:00 UTC ) this problem is not 100%
reproducable
1. vm start will get BSOD
2. vm start will work OK
3. vm start will oops kernel


----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-28 22:25

Message:
Logged In: YES 
user_id=1409122
Originator: YES

Host and Guest are 32-bit

----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-28 22:23

Message:
Logged In: YES 
user_id=1409122
Originator: YES

Windows configured with StandartPC HAL

----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-28 22:19

Message:
Logged In: YES 
user_id=1409122
Originator: YES

File Added: cpuinfo

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1881418&group_id=180599

[kvm-devel] More queries on guest swapping in KVM

[Arn <cse...@gm...>]

Hi ,

Regarding the guest swapping mechanism (host ability to swap out guest
pages),

1) Can the host swap out any page of a guest VM ? (including the guest's
kernel pages)

2) Does the guest know its page has been swapped out or is this transparent
to the guest ? ( I think the guest
wouldn't know but I just want to be sure)

3) Is there any way the guest can prevent the host from swapping out its
pages ?

thanks
Arn

Re: [kvm-devel] [kvm-ppc-devel] [PATCH] Fix kvm-userspace configure script so that cc=gcc

[Jerone Y. <jy...@us...>]

On Wed, 2008-04-30 at 16:14 -0500, Hollis Blanchard wrote:
> On Wednesday 30 April 2008 15:53:47 Jerone Young wrote:
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> > configure |    2 +-
> > 
> > 
> > This fixes compilation for cross compilers as many do not create a 
> ${cross_prefix}cc link. But the do a ${cross_prefix}gcc. This is what the 
> kernel does so this will work for everyone. This breaks some who do not have 
> a cc link (cross tools does not create), when I put a patch to remove libkvm 
> dependence on test config.mak.

The explanation came out awful :-) .  But yes the issue is that by
default cross tools does not create a ${cross_prefix}cc symlink. So you
have to say gcc.
> > 
> > Signed-off-by: Jerone Young <jy...@us...>
> > 
> > diff --git a/configure b/configure
> > --- a/configure
> > +++ b/configure
> > @@ -2,7 +2,7 @@
> > 
> >  prefix=/usr/local
> >  kerneldir=/lib/modules/$(uname -r)/build
> > -cc=cc
> > +cc=gcc
> >  ld=ld
> >  objcopy=objcopy
> >  want_module=1
> 
> To clarify: there is no such thing as "${cross_prefix}cc", so the configure 
> script is currently broken for cross-compiling.
> 

[kvm-devel] [PATCH] Fix kvm-userspace configure script so that cc=gcc

[Jerone Y. <jy...@us...>]

1 file changed, 1 insertion(+), 1 deletion(-)
configure |    2 +-


This fixes compilation for cross compilers as many do not create a ${cross_prefix}cc link. But the do a ${cross_prefix}gcc. This is what the kernel does so this will work for everyone. This breaks some who do not have a cc link (cross tools does not create), when I put a patch to remove libkvm dependence on test config.mak.

Signed-off-by: Jerone Young <jy...@us...>

diff --git a/configure b/configure
--- a/configure
+++ b/configure
@@ -2,7 +2,7 @@
 
 prefix=/usr/local
 kerneldir=/lib/modules/$(uname -r)/build
-cc=cc
+cc=gcc
 ld=ld
 objcopy=objcopy
 want_module=1

Re: [kvm-devel] [kvm-ppc-devel] [PATCH] Update kernel/Makefile and remove x86 only entries

[Hollis B. <ho...@us...>]

On Wednesday 30 April 2008 15:16:09 Avi Kivity wrote:
> >  hack = $(call _hack,$T/$(strip $1))
> > +
> > +ifneq '$(filter $(ARCH_DIR), x86)' ''
> > +HACK_FILES = kvm_main.c \
> > +             mmu.c \
> > +             vmx.c \
> > +             svm.c \
> > +             x86.c \
> > +             irq.h 
> > +endif
> >  
> >   
> 
> hack-files-x86 = ...
> hack-files-ppc = ...
> 
> hack-files = $(hack-files-$(ARCH_DIR))

Agreed; this is exactly what I had suggested previously.

-- 
Hollis Blanchard
IBM Linux Technology Center

Re: [kvm-devel] [kvm-ppc-devel] [PATCH] Fix kvm-userspace configure script so that cc=gcc

[Hollis B. <ho...@us...>]

On Wednesday 30 April 2008 15:53:47 Jerone Young wrote:
> 1 file changed, 1 insertion(+), 1 deletion(-)
> configure |    2 +-
> 
> 
> This fixes compilation for cross compilers as many do not create a 
${cross_prefix}cc link. But the do a ${cross_prefix}gcc. This is what the 
kernel does so this will work for everyone. This breaks some who do not have 
a cc link (cross tools does not create), when I put a patch to remove libkvm 
dependence on test config.mak.
> 
> Signed-off-by: Jerone Young <jy...@us...>
> 
> diff --git a/configure b/configure
> --- a/configure
> +++ b/configure
> @@ -2,7 +2,7 @@
> 
>  prefix=/usr/local
>  kerneldir=/lib/modules/$(uname -r)/build
> -cc=cc
> +cc=gcc
>  ld=ld
>  objcopy=objcopy
>  want_module=1

To clarify: there is no such thing as "${cross_prefix}cc", so the configure 
script is currently broken for cross-compiling.

-- 
Hollis Blanchard
IBM Linux Technology Center

[kvm-devel] [PATCH] [RESEND] Update kernel/Makefile and remove x86 only entries

[Jerone Y. <jy...@us...>]

1 file changed, 23 insertions(+), 13 deletions(-)
kernel/Makefile |   36 +++++++++++++++++++++++-------------


This patch removes static x86 entries and makes things work for multiple archs.

Signed-off-by: Jerone Young <jy...@us...>

diff --git a/kernel/Makefile b/kernel/Makefile
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -1,4 +1,7 @@ include ../config.mak
 include ../config.mak
+
+ARCH_DIR=$(if $(filter $(ARCH),x86_64 i386),x86,$(ARCH))
+ARCH_CONFIG=$(shell echo $(ARCH_DIR)|tr '[:lower:]' '[:upper:]') 
 
 KVERREL = $(patsubst /lib/modules/%/build,%,$(KERNELDIR))
 
@@ -18,10 +21,19 @@ _hack = mv $1 $1.orig && \
 	    | sed '/\#include/! s/\blapic\b/l_apic/g' > $1 && rm $1.orig
 
 unifdef = mv $1 $1.orig && \
-	  unifdef -DCONFIG_X86 $1.orig > $1; \
+	  unifdef -DCONFIG_$(ARCH_CONFIG) $1.orig > $1; \
           [ $$? -le 1 ] && rm $1.orig
 
 hack = $(call _hack,$T/$(strip $1))
+
+hack-files-x86 = kvm_main.c \
+		mmu.c \
+		vmx.c \
+		svm.c \
+		x86.c \
+		irq.h 
+
+hack-files = $(hack-files-$(ARCH_DIR))
 
 all::
 #	include header priority 1) $LINUX 2) $KERNELDIR 3) include-compat
@@ -49,21 +61,19 @@ header-sync:
 	rm -rf $T
 
 	rm -f include/asm
-	ln -sf asm-x86 include/asm
-	ln -sf asm-x86 include-compat/asm
+	ln -sf asm-$(ARCH_DIR) include/asm
+	ln -sf asm-$(ARCH_DIR) include-compat/asm
 
 source-sync:
 	rm -rf $T
 	rsync --exclude='*.mod.c' -R \
-             "$(LINUX)"/arch/x86/kvm/./*.[ch] \
-             "$(LINUX)"/virt/kvm/./*.[ch] \
-             $T/
-	$(call hack, kvm_main.c)
-	$(call hack, mmu.c)
-	$(call hack, vmx.c)
-	$(call hack, svm.c)
-	$(call hack, x86.c)
-	$(call hack, irq.h)
+	     "$(LINUX)"/arch/$(ARCH_DIR)/kvm/./*.[ch] \
+	     "$(LINUX)"/virt/kvm/./*.[ch] \
+	     $T/
+
+	for i in $(hack-files); \
+		do $(call hack, $$i); done
+
 	for i in $$(find $T -type f -printf '%P '); \
 		do cmp -s $$i $T/$$i || cp $T/$$i $$i; done
 	rm -rf $T
@@ -72,7 +82,7 @@ install:
 	mkdir -p $(DESTDIR)/$(INSTALLDIR)
 	cp *.ko $(DESTDIR)/$(INSTALLDIR)
 	for i in $(ORIGMODDIR)/drivers/kvm/*.ko \
-		 $(ORIGMODDIR)/arch/x86/kvm/*.ko; do \
+		 $(ORIGMODDIR)/arch/$(ARCH_DIR)/kvm/*.ko; do \
 		if [ -f "$$i" ]; then mv "$$i" "$$i.orig"; fi; \
 	done
 	/sbin/depmod -a

[kvm-devel] [PATCH] Handle vma regions with no backing page (v3)

[Anthony L. <ali...@us...>]

This patch allows VMA's that contain no backing page to be used for guest
memory.  This is a drop-in replacement for Ben-Ami's first page in his direct
mmio series.  Here, we continue to allow mmio pages to be represented in the
rmap.

Since v1, I've taken into account Andrea's suggestions at using VM_PFNMAP
instead of VM_IO and changed the BUG_ON to a return of bad_page.

Since v2, I've incorporated comments from Avi about returning bad_page instead
of NULL and fixed a typo spotted by Muli.

Signed-off-by: Anthony Liguori <ali...@us...>

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 1d7991a..9e3e4ee 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -532,6 +532,7 @@ pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn)
 	struct page *page[1];
 	unsigned long addr;
 	int npages;
+	pfn_t pfn;
 
 	might_sleep();
 
@@ -544,19 +545,38 @@ pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn)
 	npages = get_user_pages(current, current->mm, addr, 1, 1, 1, page,
 				NULL);
 
-	if (npages != 1) {
-		get_page(bad_page);
-		return page_to_pfn(bad_page);
-	}
+	if (unlikely(npages != 1)) {
+		struct vm_area_struct *vma;
 
-	return page_to_pfn(page[0]);
+		vma = find_vma(current->mm, addr);
+		if (vma == NULL || addr < vma->vm_start ||
+		    !(vma->vm_flags & VM_PFNMAP)) {
+			get_page(bad_page);
+			return page_to_pfn(bad_page);
+		}
+
+		pfn = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
+		BUG_ON(pfn_valid(pfn));
+	} else
+		pfn = page_to_pfn(page[0]);
+
+	return pfn;
 }
 
 EXPORT_SYMBOL_GPL(gfn_to_pfn);
 
 struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn)
 {
-	return pfn_to_page(gfn_to_pfn(kvm, gfn));
+	pfn_t pfn;
+
+	pfn = gfn_to_pfn(kvm, gfn);
+	if (pfn_valid(pfn))
+		return pfn_to_page(pfn);
+
+	WARN_ON(!pfn_valid(pfn));
+
+	get_page(bad_page);
+	return bad_page;
 }
 
 EXPORT_SYMBOL_GPL(gfn_to_page);
@@ -569,7 +589,8 @@ EXPORT_SYMBOL_GPL(kvm_release_page_clean);
 
 void kvm_release_pfn_clean(pfn_t pfn)
 {
-	put_page(pfn_to_page(pfn));
+	if (pfn_valid(pfn))
+		put_page(pfn_to_page(pfn));
 }
 EXPORT_SYMBOL_GPL(kvm_release_pfn_clean);
 
@@ -594,21 +615,25 @@ EXPORT_SYMBOL_GPL(kvm_set_page_dirty);
 
 void kvm_set_pfn_dirty(pfn_t pfn)
 {
-	struct page *page = pfn_to_page(pfn);
-	if (!PageReserved(page))
-		SetPageDirty(page);
+	if (pfn_valid(pfn)) {
+		struct page *page = pfn_to_page(pfn);
+		if (!PageReserved(page))
+			SetPageDirty(page);
+	}
 }
 EXPORT_SYMBOL_GPL(kvm_set_pfn_dirty);
 
 void kvm_set_pfn_accessed(pfn_t pfn)
 {
-	mark_page_accessed(pfn_to_page(pfn));
+	if (pfn_valid(pfn))
+		mark_page_accessed(pfn_to_page(pfn));
 }
 EXPORT_SYMBOL_GPL(kvm_set_pfn_accessed);
 
 void kvm_get_pfn(pfn_t pfn)
 {
-	get_page(pfn_to_page(pfn));
+	if (pfn_valid(pfn))
+		get_page(pfn_to_page(pfn));
 }
 EXPORT_SYMBOL_GPL(kvm_get_pfn);
 

Re: [kvm-devel] [PATCH 1 of 3] Remove dynamic allocation of /hypervisor node from device tree in memory

[Avi K. <av...@qu...>]

Jerone Young wrote:
> 1 file changed, 1 deletion(-)
> qemu/hw/ppc440_bamboo.c |    1 -
>
>
> In 2.6.26 wait is now enabled by default. With this the /hypervisor node will not be need to be idetified to enable the guest to go into wait state while idle.
>
> Signed-off-by: Jerone Young <jy...@us...>
>
> diff --git a/qemu/hw/ppc440_bamboo.c b/qemu/hw/ppc440_bamboo.c
> --- a/qemu/hw/ppc440_bamboo.c
> +++ b/qemu/hw/ppc440_bamboo.c
> @@ -163,7 +163,6 @@ void bamboo_init(ram_addr_t ram_size, in
>  	dt_cell(fdt, "/chosen", "linux,initrd-end",
>  				(initrd_base + initrd_size));
>  	dt_string(fdt, "/chosen", "bootargs", (char *)kernel_cmdline);
> -	dt_node(fdt, "/", "hypervisor");
>  #endif
>  
>  	if (kvm_enabled()) {
>   

I don't have this bit in my tree.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

Re: [kvm-devel] [PATCH] Update kernel/Makefile and remove x86 only entries

[Avi K. <av...@qu...>]

Jerone Young wrote:
> This patch removes static x86 entries and makes things work for multiple archs.
>
> Signed-off-by: Jerone Young <jy...@us...>
>
> diff --git a/kernel/Makefile b/kernel/Makefile
> --- a/kernel/Makefile
> +++ b/kernel/Makefile
> @@ -1,5 +1,10 @@ include ../config.mak
>  include ../config.mak
>  
> +ARCH_DIR=$(ARCH)
> +ifneq '$(filter $(ARCH_DIR), x86_64 i386)' ''
> +	ARCH_DIR=x86
> +endif
> + 
>   

Let's be consistent with functions and directives:

  ARCH_DIR=$(if $(filter $(ARCH),x86_64 i386),x86,$(ARCH))

>  KVERREL = $(patsubst /lib/modules/%/build,%,$(KERNELDIR))
>  
>  DESTDIR=
> @@ -18,10 +23,19 @@ _hack = mv $1 $1.orig && \
>  	    | sed '/\#include/! s/\blapic\b/l_apic/g' > $1 && rm $1.orig
>  
>  unifdef = mv $1 $1.orig && \
> -	  unifdef -DCONFIG_X86 $1.orig > $1; \
> +	  unifdef -DCONFIG_$(shell echo $(ARCH_DIR)|tr '[:lower:]' '[:upper:]') $1.orig > $1; \
>            [ $$? -le 1 ] && rm $1.orig
>   

My eyes hurt.  A new variable please (and assign with := to avoid the 
shell being called too often).

>  
>  hack = $(call _hack,$T/$(strip $1))
> +
> +ifneq '$(filter $(ARCH_DIR), x86)' ''
> +HACK_FILES = kvm_main.c \
> +		mmu.c \
> +		vmx.c \
> +		svm.c \
> +		x86.c \
> +		irq.h 
> +endif
>  
>   

hack-files-x86 = ...
hack-files-ppc = ...

hack-files = $(hack-files-$(ARCH_DIR))

(yeah, someone else is now gauging their eyes out)

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

[kvm-devel] [PATCH] Don't leak EPT identity page table

[Anthony L. <ali...@us...>]

In vmx.c:alloc_identity_pagetable() we grab a reference to the EPT identity
page table via gfn_to_page().  We never release this reference though.

This patch releases the reference to this page on VM destruction.  I haven't
tested this with EPT.

Signed-off-by: Anthony Liguori <ali...@us...>

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 578a0c1..63f46cf 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3909,6 +3909,8 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
 	kvm_free_physmem(kvm);
 	if (kvm->arch.apic_access_page)
 		put_page(kvm->arch.apic_access_page);
+	if (kvm->arch.ept_identity_pagetable)
+		put_page(kvm->arch.ept_identity_pagetable);
 	kfree(kvm);
 }
 

[kvm-devel] [PATCH] use common code for i386 and x86_64

[Glauber C. <gc...@re...>]

There is no reason why should i386 and x86_64 code for rdtsc be different.
Unify them.
---
 cpu-all.h |   11 +----------
 1 files changed, 1 insertions(+), 10 deletions(-)

diff --git a/cpu-all.h b/cpu-all.h
index 2a2b197..1c9e2a3 100644
--- a/cpu-all.h
+++ b/cpu-all.h
@@ -930,16 +930,7 @@ static inline int64_t cpu_get_real_ticks(void)
     return ((int64_t)h << 32) | l;
 }
 
-#elif defined(__i386__)
-
-static inline int64_t cpu_get_real_ticks(void)
-{
-    int64_t val;
-    asm volatile ("rdtsc" : "=A" (val));
-    return val;
-}
-
-#elif defined(__x86_64__)
+#elif defined(__i386__) || defined(__x86_64__)
 
 static inline int64_t cpu_get_real_ticks(void)
 {
-- 
1.5.0.6

Re: [kvm-devel] Windows PV driver for KVM

[Avi K. <av...@qu...>]

Jiang, Yunhong wrote:
> I noticed there is a windows PV driver based on virtIO in
> http://sourceforge.net/project/showfiles.php?group_id=180599
>
> But when I enable the driver in guest, the guest will hang. I'm using
> changeset around April, 18. Since the driver is created in March, I
> assume the changeset in Apri should be ok.
>
> Are there any special action needed to enable the PV driver in windows?
> Have anyone tried it recently?
>   

We are using it in production.  What HAL is the guest using?  Are you 
running with smp?

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

Re: [kvm-devel] [PATCH] libkvm: dump all errors to stderr

[Avi K. <av...@qu...>]

Jan Kiszka wrote:
> Minor cleanup I came across while reverting printf instrumentations.
>
>   
Applied, thanks.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

Re: [kvm-devel] WARNING: at /usr/src/modules/kvm/mmu.c:390 account_shadowed()

[Thomas C. <tho...@al...>]

On Sat, Apr 26, 2008 at 10:08 AM, Avi Kivity <av...@qu...> wrote:
> Marcelo Tosatti wrote:
>
> > On Wed, Apr 23, 2008 at 09:30:06AM +0300, Avi Kivity wrote:
> >
> >
> > >
> > > > as I got no reply, I guess it is a bad setup on my part. If that might
> > > > help, this happenned while I was doing a "make -j" on webkit svn tree
> > > > (ie. heavy c++ compilation workload) .
> > > >
> > > >
> > > >
> > > No this is not bad setup.  No amount of bad setup should give this
> warning.
> > >
> > > You didn't get a reply because no one knows what to make of it, and
> because it's much more fun to debate endianess or contemplete guests with
> eighty thousand disks than to fix those impossible bugs.  If you can give
> clear instructions on how to reproduce this, we will try it out.  Please be
> sure to state OS name and versions for the guest as well as the host.
> > >
> > >
> >
> > It is valid to have more than PAGES_PER_HPAGE in the largepage's
> > shadowed count. If the gpte read races with a pte-update-from-guest (and
> > the pte update results in a different sp->role), it might account twice
> > for a single gfn.
> >
> > Such "zombie" shadow pages should eventually be removed through
> > recycling, allowing for instantiation of a large page, unless references
> > can be leaked. Can't spot such leakage problem though.
> >
> >
> >
>
>  That strikes me as unlikely (though a valid scenario).
>
>  An alternative explanation is that we're seeing a nonpae guest, so each
> page can be shadowed in two different roles (two quadrants for a pte page)
> or even four (for a pgd page).
>
>  Thomas, are you running a 32-bit nonpae guest?
>

Yes it was 32bit nonpae guest on 32bit nonpae host.

Severals things :
 - I never reproduiced the problem
 - The problem happened under heavy load, and I was experiencing
slowdown on this laptop after a few hours of work. For example my
system was reporting 4500fps on glxgears after fresh boot, and 500fps
after a day of work. After some investigations, I found on the dell
website that my laptop had fans activation problems with my bios
version. After bios upgrade, my system does slows after heavy load. So
as my system was heating to much because of this bios problem, I think
you can forget my report.

Re: [kvm-devel] cirrusfb division by zero

[Anthony L. <ali...@us...>]

Marcelo Tosatti wrote:
> Anthony, 
>
> The following sequence crashes F9 guests, when using VNC:
>
> # modprobe cirrusfb
> # vbetool post
>
> Results in Floating point exception at:
>
> cirrus_do_copy()
> {
>
>     depth = s->get_bpp((VGAState *)s) / 8 
>     ...
>     sx = (src % (width * depth)) / depth;
>     ...
> }
>
> Problem is that ->get_bpp returns 0.
>
> Following band-aid "fixes it" (coff). I have no idea if its correct
> though ?
>   

It suggests something is very broken.  get_bpp only is supposed to 
return 0 when in VGA mode.  I don't think blitting should happen when in 
VGA mode.  Applying the patch is not a bad idea but what was the guest 
doing when this happened?  Was it in the process of transitioning from 
one mode to another?

Regards,

Anthony Liguori

> "vbetool post" corrupts both SDL and VNC displays when using cirrusfb,
> but seems a separate problem.
>
>
> diff --git a/qemu/hw/cirrus_vga.c b/qemu/hw/cirrus_vga.c
> index e14ec35..9f860ff 100644
> --- a/qemu/hw/cirrus_vga.c
> +++ b/qemu/hw/cirrus_vga.c
> @@ -709,6 +709,8 @@ static void cirrus_do_copy(CirrusVGAState *s, int dst, int src, int w, in     int notify = 0;
>
>      depth = s->get_bpp((VGAState *)s) / 8;
> +    if (!depth)
> +        depth = 1;
>      s->get_resolution((VGAState *)s, &width, &height);
>
>      /* extra x, y */
>
>   

Re: [kvm-devel] [PATCH 0/6] SVM: complete support for kvmtrace

[Avi K. <av...@qu...>]

Joerg Roedel wrote:
> This series of patches adds the missing kvmtrace markers to the SVM specific
> code paths in KVM. Further it adds a new trace event for TDP page faults.
>
>   

Thanks, applied all.  Please use git format-patch -n (or format.numbered 
= auto) so that patch order is explicit.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

[kvm-devel] [ kvm-Bugs-1905703 ] SMP 8-way guests fail to boot

[SourceForge.net <no...@so...>]

Bugs item #1905703, was opened at 2008-03-02 15:49
Message generated for change (Comment added) made by jroedel
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1905703&group_id=180599

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Technologov (technologov)
Assigned to: Nobody/Anonymous (nobody)
Summary: SMP 8-way guests fail to boot

Initial Comment:
Tried the recently added 8-way SMP on KVM-AMD. (Dual-CPU , Quad-Core AMD Barcelona K10)

Guests fail to boot.

Guests:
- Windows Server 2003 (32-bit)
- Windows Server 2008 (64-bit)
- Fedora 8 (64-bit)

Windows experience BSODs, while Linux just stucks during "init" phase.

While SMP 2 and 4-way are also were sometimes hanging, this one (8-way SMP hang) is consistent.

I have tried only KVM-AMD, because I don't have (yet) such an Intel machine.

-Alexey 02.03.2008.

----------------------------------------------------------------------

Comment By: Joerg Roedel (jroedel)
Date: 2008-04-30 19:53

Message:
Logged In: YES 
user_id=2019182
Originator: NO

I tried to reproduce this on an 8-way AMD box with Barcelona B3 silicon. I
was not able to reproduce the boot problems with Windows Server 2008 64
bit. Can you give more details like the qemu command line and the kvm
version you have used?

----------------------------------------------------------------------

Comment By: Marcelo Tosatti (mtosatti)
Date: 2008-04-30 18:52

Message:
Logged In: YES 
user_id=2022487
Originator: NO

Alexey,

Can you please provide more details on the problem? 

- kvm_stat --once output once the guest is hung
- is the qemu process consuming excessive CPU time on the host?
- on the Linux case, is it a hard hang or can you still see the 
guest accepting keyboard input?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1905703&group_id=180599

Re: [kvm-devel] [PATCH] kvm_show_code for ROM code

[Avi K. <av...@qu...>]

Jan Kiszka wrote:
> Userland-located ROM memory is not available via kvm->physical_memory +
> guest_address. To let kvm_show_code also dump useful information when
> some problem in ROM (BIOS...) occurs, this patch first tries to obtain
> the memory content via the mmio_read callback - maybe not 100% clean,
> but works at least for the QEMU use case. If the callback complains
> about the given address, we then fall back to RAM access.
>
>   

kvm->physical_memory is actually broken, since nothing guarantees a 1:1 
(+offset) mapping.

Why not use ->mmio_read() all the time?  Sure it overloads the 
definition of mmio_read(), but worse things have happened.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

Re: [kvm-devel] [PATCH] x86: handle double and triple faults for every exception

[Avi K. <av...@qu...>]

Jan Kiszka wrote:
> Yes, the following still works for me. But I'm not the right person to
> ask if there are obscure cases where you may not want this clearing
> while just editing some registers (I'm thinking of debugger scenarios
> now).
>
>   

Applied, as the bug we're fixing is much more likely than the bug we're 
creating.  Thanks.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

Re: [kvm-devel] KVM: PIT: support mode 4

[Avi K. <av...@qu...>]

Marcelo Tosatti wrote:
> The in-kernel PIT emulation ignores pending timers if operating under
> mode 4, which for example DragonFlyBSD uses (and Plan9 too, apparently).
>
> Mode 4 seems to be similar to one-shot mode, other than the fact that it
> starts counting after the next CLK pulse once programmed, while mode 1
> starts counting immediately, so add a FIXME to enhance precision.
>
> Fixes https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1952988&group_id=180599
>   

Applied, thanks.

-- 
Any sufficiently difficult bug is indistinguishable from a feature.

[kvm-devel] [ kvm-Bugs-1881418 ] win xp guest BSOD and host kernel oops if "-smb path"

[SourceForge.net <no...@so...>]

Bugs item #1881418, was opened at 2008-01-28 16:55
Message generated for change (Comment added) made by mtosatti
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1881418&group_id=180599

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: intel
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Alexey Fisher (fishor-flex)
Assigned to: Nobody/Anonymous (nobody)
Summary: win xp guest BSOD and host kernel oops if "-smb path"

Initial Comment:
Windows XP SP2 guest will get blueskreen and kernel ( Linux zwerg 2.6.24-rc6-g933400f4 #3 SMP PREEMPT ) will oops somtime with complete system freez  if smb share cenabled. There is no difference if "-smb /home/path" or "-smb share" used.

kvm-kernel commit version b90a99cf8e03e96a4e3347d9e9ee3750e53d07fd

kvm-userspace e82b1fc2be848acc537512cf8e00b9aea37f3fb2



----------------------------------------------------------------------

Comment By: Marcelo Tosatti (mtosatti)
Date: 2008-04-30 12:56

Message:
Logged In: YES 
user_id=2022487
Originator: NO

Can you please provide the host kernel oops? 

----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-29 13:29

Message:
Logged In: YES 
user_id=1409122
Originator: YES

with last git update ( 29.01.08 18:00 UTC ) this problem is not 100%
reproducable
1. vm start will get BSOD
2. vm start will work OK
3. vm start will oops kernel


----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-28 17:25

Message:
Logged In: YES 
user_id=1409122
Originator: YES

Host and Guest are 32-bit

----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-28 17:23

Message:
Logged In: YES 
user_id=1409122
Originator: YES

Windows configured with StandartPC HAL

----------------------------------------------------------------------

Comment By: Alexey Fisher (fishor-flex)
Date: 2008-01-28 17:19

Message:
Logged In: YES 
user_id=1409122
Originator: YES

File Added: cpuinfo

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1881418&group_id=180599

[kvm-devel] [ kvm-Bugs-1919354 ] Booting four guests likely fails

[SourceForge.net <no...@so...>]

Bugs item #1919354, was opened at 2008-03-19 04:18
Message generated for change (Comment added) made by mtosatti
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1919354&group_id=180599

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: yunfeng (yunfeng)
Assigned to: Nobody/Anonymous (nobody)
Summary: Booting four guests likely fails

Initial Comment:
Environment:
------------
Host OS:ia32e rhel5
Guest OS:ia32e rhel4u3
Commit: kernel d7f190bada15ae6c775f4ee47d1eb3f2952f62ae  
            userspace c8b1b710f3188baaffcaebfa018910f5ef43a90e
Hardware:Platform     Woodcrest
         CPU          4
         Memory size  8G'

Bug detailed description:
--------------------------
if boot four guests sequentially, the fourth guest can't boot, it will hang at
some point as the attachment shows.

Reproduce steps:
----------------
boot 4 guests using the command four times:
qemu-system-x86_64 -m 256 -monitor pty -net
nic,macaddr=00:16:3e:78:1a:5b,model=rtl8139 -net tap,script=/etc/kvm/qemu-ifup
-hda /share/xvs/var/tmp-img_CPL_VM_12_4_1205807989_1 &

Tested it for 10 times,pass rate is 70% 

Summary Test Report of Last Session
=====================================================================
                            Total   Pass    Fail    NoResult   Crash
=====================================================================
control_panel               9       6       3         0        0
Restart                     1       1       0         0        0
=====================================================================
control_panel               9       6       3         0        0
 :repeat_four_sguest_64_g   9       6       3         0        0
Restart                     1       1       0         0        0
 :Guest64_64_g64            1       1       0         0        0
=====================================================================
Total                       10      7       3         0        0


----------------------------------------------------------------------

Comment By: Marcelo Tosatti (mtosatti)
Date: 2008-04-30 12:55

Message:
Logged In: YES 
user_id=2022487
Originator: NO

This should be fixed now in kvm-67.

Yunfeng, please confirm.

----------------------------------------------------------------------

Comment By: Avi Kivity (avik)
Date: 2008-03-24 09:15

Message:
Logged In: YES 
user_id=539971
Originator: NO

I see something similar with FC6, but the guest continues after a few
seconds.

Will try to reproduce with centos.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1919354&group_id=180599

[kvm-devel] [ kvm-Bugs-1905703 ] SMP 8-way guests fail to boot

[SourceForge.net <no...@so...>]

Bugs item #1905703, was opened at 2008-03-02 10:49
Message generated for change (Comment added) made by mtosatti
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1905703&group_id=180599

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Technologov (technologov)
Assigned to: Nobody/Anonymous (nobody)
Summary: SMP 8-way guests fail to boot

Initial Comment:
Tried the recently added 8-way SMP on KVM-AMD. (Dual-CPU , Quad-Core AMD Barcelona K10)

Guests fail to boot.

Guests:
- Windows Server 2003 (32-bit)
- Windows Server 2008 (64-bit)
- Fedora 8 (64-bit)

Windows experience BSODs, while Linux just stucks during "init" phase.

While SMP 2 and 4-way are also were sometimes hanging, this one (8-way SMP hang) is consistent.

I have tried only KVM-AMD, because I don't have (yet) such an Intel machine.

-Alexey 02.03.2008.

----------------------------------------------------------------------

Comment By: Marcelo Tosatti (mtosatti)
Date: 2008-04-30 12:52

Message:
Logged In: YES 
user_id=2022487
Originator: NO

Alexey,

Can you please provide more details on the problem? 

- kvm_stat --once output once the guest is hung
- is the qemu process consuming excessive CPU time on the host?
- on the Linux case, is it a hard hang or can you still see the 
guest accepting keyboard input?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1905703&group_id=180599