On Wed, Feb 26, 2014 at 06:01:53PM -0600, Alex Thorlton wrote:
> On Wed, Feb 26, 2014 at 03:24:34PM -0800, Andrew Morton wrote:
> > On Wed, 26 Feb 2014 13:39:17 -0600 Alex Thorlton wrote:
> >
> > > >
> > > >
> > > > > I'd suggest the patch below on top of your changes, but I won't argue.
> > >
On 02/26, Gerald Schaefer wrote:
>
> On Wed, 26 Feb 2014 19:06:03 +0100
> Oleg Nesterov wrote:
>
> > It would be nice to also change thp_split_mm() to not not play with
> > mm->def_flags, but I am not sure if we can do this.
>
> Hmm, I'm also wondering about this. Basically, we only need
On 02/26, Paolo Bonzini wrote:
>
> > +#ifdef CONFIG_S390
> > + /*
> > +* MADV_HUGEPAGE is broken after s390_enable_sie(), qemu
> > +* blindly does madvise(MADV_HUGEPAGE) for for all kvm pages
> > +* and expects it must fail on s390. Avoid a possible
On 02/26, Paolo Bonzini wrote:
+#ifdef CONFIG_S390
+ /*
+* MADV_HUGEPAGE is broken after s390_enable_sie(), qemu
+* blindly does madvise(MADV_HUGEPAGE) for for all kvm pages
+* and expects it must fail on s390. Avoid a possible SIGSEGV
+
On 02/26, Gerald Schaefer wrote:
On Wed, 26 Feb 2014 19:06:03 +0100
Oleg Nesterov o...@redhat.com wrote:
It would be nice to also change thp_split_mm() to not not play with
mm-def_flags, but I am not sure if we can do this.
Hmm, I'm also wondering about this. Basically, we only need
On Wed, Feb 26, 2014 at 06:01:53PM -0600, Alex Thorlton wrote:
On Wed, Feb 26, 2014 at 03:24:34PM -0800, Andrew Morton wrote:
On Wed, 26 Feb 2014 13:39:17 -0600 Alex Thorlton athorl...@sgi.com wrote:
I'd suggest the patch below on top of your changes, but I won't argue.
On Wed, Feb 26, 2014 at 03:24:34PM -0800, Andrew Morton wrote:
> On Wed, 26 Feb 2014 13:39:17 -0600 Alex Thorlton wrote:
>
> > >
> > >
> > > > I'd suggest the patch below on top of your changes, but I won't argue.
> >
> > I like this approach, and your updated comment as well. This should
>
On Wed, 26 Feb 2014 13:39:17 -0600 Alex Thorlton wrote:
> >
> >
> > > I'd suggest the patch below on top of your changes, but I won't argue.
>
> I like this approach, and your updated comment as well. This should
> probably go in as [PATCH 2/4] in my series. Do I need to spin a v5
> with
> +#ifdef CONFIG_S390
> + /*
> + * MADV_HUGEPAGE is broken after s390_enable_sie(), qemu
> + * blindly does madvise(MADV_HUGEPAGE) for for all kvm pages
> + * and expects it must fail on s390. Avoid a possible SIGSEGV
> + * until
On Wed, Feb 26, 2014 at 08:27:36PM +0100, Christian Borntraeger wrote:
> On 26/02/14 19:06, Oleg Nesterov wrote:
> > On 02/26, Alex Thorlton wrote:
> >>
> >> + * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
> >> + * qemu blindly does madvise(MADV_HUGEPAGE) after
On 26/02/14 19:06, Oleg Nesterov wrote:
> On 02/26, Alex Thorlton wrote:
>>
>> + * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
>> + * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
>> + */
>> +#ifdef CONFIG_S390
>> + if
On Wed, 26 Feb 2014 19:06:03 +0100
Oleg Nesterov wrote:
> On 02/26, Alex Thorlton wrote:
> >
> > + * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
> > + * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
> > + */
> > +#ifdef CONFIG_S390
> > + if
On 02/26, Peter Zijlstra wrote:
>
> On Wed, Feb 26, 2014 at 04:31:44PM +0100, Oleg Nesterov wrote:
> > /*
> > @@ -1977,8 +1975,12 @@ int hugepage_madvise(struct vm_area_stru
> > */
> > if (*vm_flags & (VM_HUGEPAGE | VM_NO_THP))
> > return
On 02/26, Alex Thorlton wrote:
>
> + * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
> + * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
> + */
> +#ifdef CONFIG_S390
> + if (mm_has_pgste(vma->vm_mm))
> return -EINVAL;
>
On Wed, Feb 26, 2014 at 05:57:59PM +0100, Peter Zijlstra wrote:
> On Wed, Feb 26, 2014 at 04:31:44PM +0100, Oleg Nesterov wrote:
> Do we want a comment here, explaining why s390 is special again?
Here's what I've got, with everybody's suggestions spun together:
diff --git a/mm/huge_memory.c
On Wed, Feb 26, 2014 at 04:31:44PM +0100, Oleg Nesterov wrote:
> --- x/arch/s390/mm/pgtable.c
> +++ x/arch/s390/mm/pgtable.c
> @@ -1084,7 +1084,6 @@ static inline void thp_split_mm(struct m
> vma->vm_flags &= ~VM_HUGEPAGE;
> vma->vm_flags |= VM_NOHUGEPAGE;
> }
> -
On Wed, 26 Feb 2014 16:31:44 +0100
Oleg Nesterov wrote:
> On 02/26, Christian Borntraeger wrote:
> >
> > On 26/02/14 15:50, Oleg Nesterov wrote:
> > >
> > > But perhaps qemu can be changed to avoid MADV_HUGEPAGE on s390 ?
> > > Otherwise I'd suggest the change below.
> > >
> > > Oleg.
> > >
> >
On 02/26, Christian Borntraeger wrote:
>
> On 26/02/14 15:50, Oleg Nesterov wrote:
> >
> > But perhaps qemu can be changed to avoid MADV_HUGEPAGE on s390 ?
> > Otherwise I'd suggest the change below.
> >
> > Oleg.
> >
> >
> > --- x/mm/huge_memory.c
> > +++ x/mm/huge_memory.c
> > @@ -1968,8 +1968,6
Christian Borntraeger wrote:
> On 26/02/14 15:50, Oleg Nesterov wrote:
> [...]
>
> >> NAK.
> >>
> >> Since 2012 qemu does call "qemu_madvise(new_block->host, size,
> >> QEMU_MADV_HUGEPAGE);" for all kvm pages.
> >> (commit ad0b5321f1f797274603ebbe20108b0750baee94 Call MADV_HUGEPAGE for
> >>
On 26/02/14 15:50, Oleg Nesterov wrote:
[...]
>> NAK.
>>
>> Since 2012 qemu does call "qemu_madvise(new_block->host, size,
>> QEMU_MADV_HUGEPAGE);" for all kvm pages.
>> (commit ad0b5321f1f797274603ebbe20108b0750baee94 Call MADV_HUGEPAGE for
>> guest RAM allocations) so this
>> breaks any
On 02/26, Christian Borntraeger wrote:
>
> On 26/02/14 00:53, a...@linux-foundation.org wrote:
> > Subject: + mm-revert-thp-make-madv_hugepage-check-for-mm-def_flags.patch
> > added to -mm tree
> > To:
> > athorl...@sgi.com,aarca...@redhat.com,borntrae...@de
On 02/26, Christian Borntraeger wrote:
On 26/02/14 00:53, a...@linux-foundation.org wrote:
Subject: + mm-revert-thp-make-madv_hugepage-check-for-mm-def_flags.patch
added to -mm tree
To:
athorl...@sgi.com,aarca...@redhat.com,borntrae...@de.ibm.com,ebied...@xmission.com,gerald.schae
On 26/02/14 15:50, Oleg Nesterov wrote:
[...]
NAK.
Since 2012 qemu does call qemu_madvise(new_block-host, size,
QEMU_MADV_HUGEPAGE); for all kvm pages.
(commit ad0b5321f1f797274603ebbe20108b0750baee94 Call MADV_HUGEPAGE for
guest RAM allocations) so this
breaks any recent kvm guest on
Christian Borntraeger wrote:
On 26/02/14 15:50, Oleg Nesterov wrote:
[...]
NAK.
Since 2012 qemu does call qemu_madvise(new_block-host, size,
QEMU_MADV_HUGEPAGE); for all kvm pages.
(commit ad0b5321f1f797274603ebbe20108b0750baee94 Call MADV_HUGEPAGE for
guest RAM allocations) so
On 02/26, Christian Borntraeger wrote:
On 26/02/14 15:50, Oleg Nesterov wrote:
But perhaps qemu can be changed to avoid MADV_HUGEPAGE on s390 ?
Otherwise I'd suggest the change below.
Oleg.
--- x/mm/huge_memory.c
+++ x/mm/huge_memory.c
@@ -1968,8 +1968,6 @@ out:
int
On Wed, 26 Feb 2014 16:31:44 +0100
Oleg Nesterov o...@redhat.com wrote:
On 02/26, Christian Borntraeger wrote:
On 26/02/14 15:50, Oleg Nesterov wrote:
But perhaps qemu can be changed to avoid MADV_HUGEPAGE on s390 ?
Otherwise I'd suggest the change below.
Oleg.
---
On Wed, Feb 26, 2014 at 04:31:44PM +0100, Oleg Nesterov wrote:
--- x/arch/s390/mm/pgtable.c
+++ x/arch/s390/mm/pgtable.c
@@ -1084,7 +1084,6 @@ static inline void thp_split_mm(struct m
vma-vm_flags = ~VM_HUGEPAGE;
vma-vm_flags |= VM_NOHUGEPAGE;
}
-
On Wed, Feb 26, 2014 at 05:57:59PM +0100, Peter Zijlstra wrote:
On Wed, Feb 26, 2014 at 04:31:44PM +0100, Oleg Nesterov wrote:
Do we want a comment here, explaining why s390 is special again?
Here's what I've got, with everybody's suggestions spun together:
diff --git a/mm/huge_memory.c
On 02/26, Alex Thorlton wrote:
+ * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
+ * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
+ */
+#ifdef CONFIG_S390
+ if (mm_has_pgste(vma-vm_mm))
return -EINVAL;
+#endif
On 02/26, Peter Zijlstra wrote:
On Wed, Feb 26, 2014 at 04:31:44PM +0100, Oleg Nesterov wrote:
/*
@@ -1977,8 +1975,12 @@ int hugepage_madvise(struct vm_area_stru
*/
if (*vm_flags (VM_HUGEPAGE | VM_NO_THP))
return -EINVAL;
-
On Wed, 26 Feb 2014 19:06:03 +0100
Oleg Nesterov o...@redhat.com wrote:
On 02/26, Alex Thorlton wrote:
+ * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
+ * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
+ */
+#ifdef CONFIG_S390
+ if
On 26/02/14 19:06, Oleg Nesterov wrote:
On 02/26, Alex Thorlton wrote:
+ * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
+ * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
+ */
+#ifdef CONFIG_S390
+ if (mm_has_pgste(vma-vm_mm))
On Wed, Feb 26, 2014 at 08:27:36PM +0100, Christian Borntraeger wrote:
On 26/02/14 19:06, Oleg Nesterov wrote:
On 02/26, Alex Thorlton wrote:
+ * MADV_HUGEPAGE after PRCTL_THP_DISABLE is broken on s390 because
+ * qemu blindly does madvise(MADV_HUGEPAGE) after s390_enable_sie().
+ */
+#ifdef CONFIG_S390
+ /*
+ * MADV_HUGEPAGE is broken after s390_enable_sie(), qemu
+ * blindly does madvise(MADV_HUGEPAGE) for for all kvm pages
+ * and expects it must fail on s390. Avoid a possible SIGSEGV
+ * until qemu is
On Wed, 26 Feb 2014 13:39:17 -0600 Alex Thorlton athorl...@sgi.com wrote:
I'd suggest the patch below on top of your changes, but I won't argue.
I like this approach, and your updated comment as well. This should
probably go in as [PATCH 2/4] in my series. Do I need to spin a v5
On Wed, Feb 26, 2014 at 03:24:34PM -0800, Andrew Morton wrote:
On Wed, 26 Feb 2014 13:39:17 -0600 Alex Thorlton athorl...@sgi.com wrote:
I'd suggest the patch below on top of your changes, but I won't argue.
I like this approach, and your updated comment as well. This should
36 matches
Mail list logo
