* Re: [SeaBIOS] IO APIC emulation failure with qemu-kvm
From: Avi Kivity @ 2011-02-07 9:54 UTC (permalink / raw)
To: Ravi Kumar Kulkarni; +Cc: Jan Kiszka, seabios, qemu-devel, kvm, virtualization
In-Reply-To: <AANLkTi=p1R63Oaw9BkZ_YLVB54Yi7mQkFt5L6W5Um1ph@mail.gmail.com>
On 02/07/2011 11:47 AM, Ravi Kumar Kulkarni wrote:
> >
> > That is not the same address. And the code you posted doesn't make any
> > sense.
> >
> sorry for the mistake. here's the correct one
>
>
> (qemu) xp /20iw 0x1e2f3f7b
> 0x000000001e2f3f7b: (bad)
> 0x000000001e2f3f7c: std
> 0x000000001e2f3f7d: (bad)
> 0x000000001e2f3f7e: (bad)
That looks like garbage. Are you sure you're disassembling the right code?
--
error compiling committee.c: too many arguments to function
^ permalink raw reply
* Re: [SeaBIOS] IO APIC emulation failure with qemu-kvm
From: Ravi Kumar Kulkarni @ 2011-02-07 10:28 UTC (permalink / raw)
To: Avi Kivity; +Cc: Jan Kiszka, seabios, qemu-devel, kvm, virtualization
In-Reply-To: <4D4FC13B.7010001@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 2034 bytes --]
On Mon, Feb 7, 2011 at 3:24 PM, Avi Kivity <avi@redhat.com> wrote:
> On 02/07/2011 11:47 AM, Ravi Kumar Kulkarni wrote:
>>
>> >
>> > That is not the same address. And the code you posted doesn't make any
>> > sense.
>> >
>> sorry for the mistake. here's the correct one
>>
>>
>> (qemu) xp /20iw 0x1e2f3f7b
>> 0x000000001e2f3f7b: (bad)
>> 0x000000001e2f3f7c: std
>> 0x000000001e2f3f7d: (bad)
>> 0x000000001e2f3f7e: (bad)
>
> That looks like garbage. Are you sure you're disassembling the right code?
>
ok . Just to be clear i ran the command qemu-kvm once and i found
got the crash report below which i have attached and in that eip is at
0x1e2f3f77
and then
(qemu) xp /20iw 0x1e2f3f77
0x000000001e2f3f77: pop %ds
0x000000001e2f3f78: inc %edx
0x000000001e2f3f79: loope 0x1e2f3fc8
0x000000001e2f3f7b: pop %ds
0x000000001e2f3f7c: jnp 0x1e2f3f5e
0x000000001e2f3f7e: dec %ebp
0x000000001e2f3f7f: pop %ds
0x000000001e2f3f80: xchg %eax,%esp
0x000000001e2f3f81: aas
0x000000001e2f3f82: das
0x000000001e2f3f83: push %ds
0x000000001e2f3f84: add %eax,(%eax)
0x000000001e2f3f86: add %al,(%eax)
0x000000001e2f3f88: add %al,(%eax)
0x000000001e2f3f8a: add %al,(%eax)
0x000000001e2f3f8c: add %al,(%eax)
0x000000001e2f3f8e: add %al,(%eax)
0x000000001e2f3f90: or $0x0,%eax
0x000000001e2f3f95: add %al,(%eax)
0x000000001e2f3f97: add %cl,0x81e2f4c(%esi)
Warm Regards,
Ravi Kulkarni.
>
[-- Attachment #2: kvm1.txt --]
[-- Type: text/plain, Size: 2128 bytes --]
KVM internal error. Suberror: 1
rax 000000000000000d rbx 0000000000000001 rcx 00000000ffffffff rdx 0000000000000001
rsi 000000001f4de142 rdi 000000001f4de0e2 rsp 00000000000c0004 rbp 000000001f4de07b
r8 0000000000000000 r9 0000000000000000 r10 0000000000000000 r11 0000000000000000
r12 0000000000000000 r13 0000000000000000 r14 0000000000000000 r15 0000000000000000
rip 000000001e2f3f77 rflags 00010002
cs 0008 (00000000/ffffffff p 1 dpl 0 db 1 s 1 type b l 0 g 1 avl 0)
ds 0003 (00000000/ffffffff p 0 dpl 0 db 0 s 0 type 0 l 0 g 0 avl 0)
es 0040 (00000000/ffffffff p 1 dpl 3 db 1 s 1 type 3 l 0 g 1 avl 0)
ss 0010 (00000000/ffffffff p 1 dpl 0 db 1 s 1 type 3 l 0 g 1 avl 0)
fs 0040 (00000000/ffffffff p 1 dpl 3 db 1 s 1 type 3 l 0 g 1 avl 0)
gs 0040 (00000000/ffffffff p 1 dpl 3 db 1 s 1 type 3 l 0 g 1 avl 0)
tr 004b (1f522206/00000067 p 1 dpl 3 db 0 s 0 type b l 0 g 0 avl 1)
ldt 0000 (00000000/0000ffff p 1 dpl 0 db 0 s 0 type 2 l 0 g 0 avl 0)
gdt 1f522006/7f
idt 1f9af000/7ff
cr0 11 cr2 0 cr3 0 cr4 0 cr8 0 efer 0
emulation failure, check dmesg for details
[-- Attachment #3: Type: text/plain, Size: 184 bytes --]
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/virtualization
^ permalink raw reply
* Re: [SeaBIOS] IO APIC emulation failure with qemu-kvm
From: Avi Kivity @ 2011-02-07 13:17 UTC (permalink / raw)
To: Ravi Kumar Kulkarni; +Cc: Jan Kiszka, seabios, qemu-devel, kvm, virtualization
In-Reply-To: <AANLkTi=Neds4Zm8sVE4k8yA_YQJ+TqMhTSRZEMKZcPhs@mail.gmail.com>
On 02/07/2011 12:28 PM, Ravi Kumar Kulkarni wrote:
> On Mon, Feb 7, 2011 at 3:24 PM, Avi Kivity<avi@redhat.com> wrote:
> > On 02/07/2011 11:47 AM, Ravi Kumar Kulkarni wrote:
> >>
> >> >
> >> > That is not the same address. And the code you posted doesn't make any
> >> > sense.
> >> >
> >> sorry for the mistake. here's the correct one
> >>
> >>
> >> (qemu) xp /20iw 0x1e2f3f7b
> >> 0x000000001e2f3f7b: (bad)
> >> 0x000000001e2f3f7c: std
> >> 0x000000001e2f3f7d: (bad)
> >> 0x000000001e2f3f7e: (bad)
> >
> > That looks like garbage. Are you sure you're disassembling the right code?
> >
> ok . Just to be clear i ran the command qemu-kvm once and i found
> got the crash report below which i have attached and in that eip is at
> 0x1e2f3f77
> and then
>
> (qemu) xp /20iw 0x1e2f3f77
> 0x000000001e2f3f77: pop %ds
> 0x000000001e2f3f78: inc %edx
> 0x000000001e2f3f79: loope 0x1e2f3fc8
> 0x000000001e2f3f7b: pop %ds
> 0x000000001e2f3f7c: jnp 0x1e2f3f5e
> 0x000000001e2f3f7e: dec %ebp
> 0x000000001e2f3f7f: pop %ds
> 0x000000001e2f3f80: xchg %eax,%esp
> 0x000000001e2f3f81: aas
> 0x000000001e2f3f82: das
This still doesn't look like real code. The problem was likely much
earlier and caused a branch into a data section.
Someone with a good understanding of your OS needs to examine the trace
and see what went wrong.
--
error compiling committee.c: too many arguments to function
^ permalink raw reply
* Re: [PATCH] sched: provide scheduler_ipi() callback in response to smp_send_reschedule()
From: Peter Zijlstra @ 2011-02-07 13:54 UTC (permalink / raw)
To: Benjamin Herrenschmidt
Cc: Richard Henderson, Ivan Kokshaysky, Matt Turner, Russell King,
Mike Frysinger, Mikael Starvik, Jesper Nilsson, Tony Luck,
Fenghua Yu, Hirokazu Takata, Ralf Baechle, David Howells,
Koichi Yasutake, Kyle McMartin, Helge Deller,
James E.J. Bottomley, Paul Mackerras, Martin Schwidefsky,
Heiko Carstens, linux390
In-Reply-To: <1297034792.14982.10.camel@pasglop>
On Mon, 2011-02-07 at 10:26 +1100, Benjamin Herrenschmidt wrote:
> You missed:
>
> diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
> index 9813605..467d122 100644
> --- a/arch/powerpc/kernel/smp.c
> +++ b/arch/powerpc/kernel/smp.c
> @@ -98,6 +98,7 @@ void smp_message_recv(int msg)
> break;
> case PPC_MSG_RESCHEDULE:
> /* we notice need_resched on exit */
> + scheduler_ipi();
> break;
> case PPC_MSG_CALL_FUNC_SINGLE:
> generic_smp_call_function_single_interrupt();
>
> Fold that in and add:
>
> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Thanks Ben!
^ permalink raw reply
* ICAC2011 Call for Posters and Workshop Papers
From: Ming Zhao @ 2011-02-08 1:11 UTC (permalink / raw)
To: virtualization
--------------------------------------------------------------------
8th International Conference on Autonomic Computing Call for Posters
--------------------------------------------------------------------
Posters are an excellent opportunity for presenting innovative research
ideas, projects and results. Accepted papers for posters (2 page) will
be published in proceedings by ACM, which will be distributed at the
conference. Posters will be displayed at the conference.
Topics of interest:
* Applications of autonomic computing
* Autonomic computing components and services
* Algorithms, theory and foundations of autonomic computing
Submission deadline: March 25 2011
More information: http://icac2011.cs.fiu.edu/workshop#poster
----------------------------------------------------------
1st IEEE/ACM Workshop on Autonomic Computing for Economics
----------------------------------------------------------
This workshop is intended as a venue for researchers interested in
economic models for autonomic systems, the use of autonomic computing to
simulate and/or create novel economic models and mechanisms and where
autonomic capabilities play a critical role in (business) applications
that fuel or require economically-driven decisions.
Important Dates:
* Paper submission: February 21, 2011
* Author notification: March 14, 2011
* Camera-ready final paper: April 2011
More Information: http://www.im.uni-karlsruhe.de/ICAC-ACE
------------------------------------------------------------------
3rd Workshop on Bio-Inspired and Self-* Algorithms for Distributed
Systems
------------------------------------------------------------------
This full day workshop aims to gather scientists, engineers, and
practitioners to share and exchange their experiences, discuss
challenges, and report state-of-the-art and in-progress research on
bio-inspired algorithms and systems.
Important Dates:
* Paper submission: February 21, 2011
* Author notification: March 2011
* Camera-ready final paper: April 2011
More Information: http://bads.icar.cnr.it/
-----------------------------------------------------------------
6th International Workshop on Feedback Control Implementation and
Design in Computing Systems and Networks
-----------------------------------------------------------------
The workshop aims to bring together researchers, practitioners, system
administrators, system programmers, and others interested in the latest
advances in the control of computer systems and networks. The focus of
the workshop is on novel, practical, systems-oriented work.
Important Dates:
* Paper submission: February 21, 2011
* Author notification: March 14, 2011
* Camera-ready final paper: April 15, 2011
More Information: http://controlofsystems.org/febid2011/
---------------------------------------------
First International E-Energy Market Challenge
---------------------------------------------
The worksohp asks for submission of agent-based models and corresponding
papers describing the design of trading agents for retail energy
markets.
Important Dates:
* Paper submission: February 28, 2011
* Author notification: March 15, 2011
* Camera-ready final paper: April 15, 2011
More Information: http://controlofsystems.org/febid2011/
-----------------------------
Workshop on Organic Computing
-----------------------------
This workshop will be a platform for the presentation of results of the
SPP projects as well as projects from the Autonomic Computing community.
These presentations will be the basis for a panel discussion on the
future directions of OC. The workshop is open for contributions with
OC-related topics from inside and outside the SPP.
Important Dates:
* Paper submission: February 21, 2011
* Author notification: March 14, 2011
* Camera-ready final paper: April 15, 2011
More Information: http://www.sra.uni-hannover.de/orgcomp/news/date/2010/12/07/workshop-organic-computing
^ permalink raw reply
* Re: [PATCH] sched: provide scheduler_ipi() callback in response to smp_send_reschedule()
From: Benjamin Herrenschmidt @ 2011-02-09 6:14 UTC (permalink / raw)
To: Peter Zijlstra
Cc: linux-m32r-ja-rQhvJZKUsGBRYuoOT4C5/9i2O/JbrIOy,
linux-mips-6z/3iImG2C8G8FEW9MqTrA,
linux-m32r-rQhvJZKUsGBRYuoOT4C5/9i2O/JbrIOy,
linux-ia64-u79uwXL29TY76Z2rM5mHXA,
linux-sh-u79uwXL29TY76Z2rM5mHXA, Heiko Carstens, Howells,
Paul Mackerras, H. Peter Anvin, sparclinux-u79uwXL29TY76Z2rM5mHXA,
Linux-Arch, linux-s390-u79uwXL29TY76Z2rM5mHXA, Jesper Nilsson,
Mikael-ZG0+EudsQA8dtHy/vicBwGD2FQJk+8+b, Russell King, Takata,
x86-DgEjT+Ai2ygdnm+yROfE0A, James E.J. Bottomley,
virtualization-qjLDD68F18O7TbgM5vRIOg, Ingo Molnar,
xen-devel-GuqFBffKawuULHF6PoxzQEEOCMrvLtNR,
Chris-ZG0+EudsQA8dtHy/vicBwGD2FQJk+8+b, Matt Turner,
uclinux-dist-devel-ZG0+EudsQA8dtHy/vicBwGD2FQJk+8+b, Fen
In-Reply-To: <1297086859.13327.16.camel@laptop>
On Mon, 2011-02-07 at 14:54 +0100, Peter Zijlstra wrote:
> On Mon, 2011-02-07 at 10:26 +1100, Benjamin Herrenschmidt wrote:
> > You missed:
> >
> > diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
> > index 9813605..467d122 100644
> > --- a/arch/powerpc/kernel/smp.c
> > +++ b/arch/powerpc/kernel/smp.c
> > @@ -98,6 +98,7 @@ void smp_message_recv(int msg)
> > break;
> > case PPC_MSG_RESCHEDULE:
> > /* we notice need_resched on exit */
> > + scheduler_ipi();
> > break;
> > case PPC_MSG_CALL_FUNC_SINGLE:
> > generic_smp_call_function_single_interrupt();
> >
> > Fold that in and add:
> >
> > Acked-by: Benjamin Herrenschmidt <benh-XVmvHMARGAS8U2dJNN8I7kB+6BGkLq7r@public.gmane.org>
>
> Thanks Ben!
BTW. I we lurking at some of our CPU hotplug code and I think I cannot
totally guarantee that this won't be called on an offline CPU. If that's
a problem, you may want to add a test for that.
IE. The call function IPIs are normally not going to be sent to an
offlined CPU, and stop_machine should be a good enough fence here, but
we do abuse reschedule for a number of things (including in some case
to wake up a sleeping CPU that was pseudo-offlined :-)
Cheers,
Ben.
^ permalink raw reply
* Re: [PATCH] sched: provide scheduler_ipi() callback in response to smp_send_reschedule()
From: Peter Zijlstra @ 2011-02-09 9:37 UTC (permalink / raw)
To: Benjamin Herrenschmidt
Cc: Richard Henderson, Ivan Kokshaysky, Matt Turner, Russell King,
Mike Frysinger, Mikael Starvik, Jesper Nilsson, Tony Luck,
Fenghua Yu, Hirokazu Takata, Ralf Baechle, David Howells,
Koichi Yasutake, Kyle McMartin, Helge Deller,
James E.J. Bottomley, Paul Mackerras, Martin Schwidefsky,
Heiko Carstens, linux390
In-Reply-To: <1297232054.14982.346.camel@pasglop>
On Wed, 2011-02-09 at 17:14 +1100, Benjamin Herrenschmidt wrote:
> On Mon, 2011-02-07 at 14:54 +0100, Peter Zijlstra wrote:
> > On Mon, 2011-02-07 at 10:26 +1100, Benjamin Herrenschmidt wrote:
> > > You missed:
> > >
> > > diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
> > > index 9813605..467d122 100644
> > > --- a/arch/powerpc/kernel/smp.c
> > > +++ b/arch/powerpc/kernel/smp.c
> > > @@ -98,6 +98,7 @@ void smp_message_recv(int msg)
> > > break;
> > > case PPC_MSG_RESCHEDULE:
> > > /* we notice need_resched on exit */
> > > + scheduler_ipi();
> > > break;
> > > case PPC_MSG_CALL_FUNC_SINGLE:
> > > generic_smp_call_function_single_interrupt();
> > >
> > > Fold that in and add:
> > >
> > > Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> >
> > Thanks Ben!
>
> BTW. I we lurking at some of our CPU hotplug code and I think I cannot
> totally guarantee that this won't be called on an offline CPU. If that's
> a problem, you may want to add a test for that.
>
> IE. The call function IPIs are normally not going to be sent to an
> offlined CPU, and stop_machine should be a good enough fence here, but
> we do abuse reschedule for a number of things (including in some case
> to wake up a sleeping CPU that was pseudo-offlined :-)
Hmm, I _think_ that should all work out nicely, but we'll see, if when
this stuff hits the tree powerpc machines start falling over we'd better
put that check in ;-)
Meanwhile I'm going to preserve this comment in the changelog of this
patch so we don't forget.
^ permalink raw reply
* [PATCH] virtio-net: add schedule check to napi_enable call in refill_work
From: Ken Stailey @ 2011-02-09 20:29 UTC (permalink / raw)
To: virtualization
Justification:
Impact: Under heavy network I/O load virtio-net driver crashes making VM guest unusable.
Testcases:
1) Sergey Svishchev reports that servers that run java webapps with high java heap usage (especially when heap size is close to physical memory size) helps trigger one of aforementioned bugs. Unfortunately, I don't have a simple test case.
2) Peter Lieven reports that his binary NNTP newsfeed test servers crash without this patch.
3) Bruce Rogers of Novell has asked for this patch to be integrated but the request was mysteriously ignored. Is is purported that this patch is being distributed with SLES.
4) I can crash 2.6.32 and 2.6.38-rc4 by simply running "scp -r /nfs/read-only/1 otherhost:/target/1" and "scp -r /nfs/read-only/2 otherhost:/target/2" concurrently with a mix of small to medium files for a few hours usually. I've never seen more than 200 GB copied before the crash occurs. Both 2.6.32 and 2.6.38-rc4 with this patch will copy more than 200GB unfailingly this way.
See
https://bugs.launchpad.net/bugs/579276
for more details.
--- drivers/net/virtio_net.c.orig 2011-02-08 14:34:51.444099190 -0500
+++ drivers/net/virtio_net.c 2011-02-08 14:18:00.484400134 -0500
@@ -446,6 +446,20 @@
}
}
+static void virtnet_napi_enable(struct virtnet_info *vi)
+{
+ napi_enable(&vi->napi);
+
+ /* If all buffers were filled by other side before we napi_enabled, we
+ * won't get another interrupt, so process any outstanding packets
+ * now. virtnet_poll wants re-enable the queue, so we disable here.
+ * We synchronize against interrupts via NAPI_STATE_SCHED */
+ if (napi_schedule_prep(&vi->napi)) {
+ virtqueue_disable_cb(vi->rvq);
+ __napi_schedule(&vi->napi);
+ }
+}
+
static void refill_work(struct work_struct *work)
{
struct virtnet_info *vi;
@@ -454,7 +468,7 @@
vi = container_of(work, struct virtnet_info, refill.work);
napi_disable(&vi->napi);
still_empty = !try_fill_recv(vi, GFP_KERNEL);
- napi_enable(&vi->napi);
+ virtnet_napi_enable(vi);
/* In theory, this can happen: if we don't get any buffers in
* we will *never* try to fill again. */
@@ -638,16 +652,7 @@
{
struct virtnet_info *vi = netdev_priv(dev);
- napi_enable(&vi->napi);
-
- /* If all buffers were filled by other side before we napi_enabled, we
- * won't get another interrupt, so process any outstanding packets
- * now. virtnet_poll wants re-enable the queue, so we disable here.
- * We synchronize against interrupts via NAPI_STATE_SCHED */
- if (napi_schedule_prep(&vi->napi)) {
- virtqueue_disable_cb(vi->rvq);
- __napi_schedule(&vi->napi);
- }
+ virtnet_napi_enable(vi);
return 0;
}
^ permalink raw reply
* Re: [PATCH] virtio-net: add schedule check to napi_enable call in refill_work
From: Rusty Russell @ 2011-02-10 1:31 UTC (permalink / raw)
To: virtualization; +Cc: Bruce Rogers, netdev
In-Reply-To: <132885.30568.qm@web110313.mail.gq1.yahoo.com>
On Thu, 10 Feb 2011 06:59:25 am Ken Stailey wrote:
> Justification:
>
> Impact: Under heavy network I/O load virtio-net driver crashes making VM guest unusable.
Hmm, this went badly wrong. I acked this patch, and it was mailed to
netdev six months ago.
Bruce's patch used spaces instead of tabs, but that should not have caused
it to be dropped. I've taken that and ported it forwards, will repost now.
Thanks for picking this up off the floor!
Rusty.
^ permalink raw reply
* [PATCH] virtio_net: Add schedule check to napi_enable call
From: Rusty Russell @ 2011-02-10 2:02 UTC (permalink / raw)
To: Herbert Xu; +Cc: netdev, David Miller, virtualization
From: "Bruce Rogers" <brogers@novell.com>
Under harsh testing conditions, including low memory, the guest would
stop receiving packets. With this patch applied we no longer see any
problems in the driver while performing these tests for extended periods
of time.
Make sure napi is scheduled subsequent to each napi_enable.
Signed-off-by: Bruce Rogers <brogers@novell.com>
Signed-off-by: Olaf Kirch <okir@suse.de>
Cc: stable@kernel.org
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
---
drivers/net/virtio_net.c | 27 ++++++++++++++++-----------
1 file changed, 16 insertions(+), 11 deletions(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -446,6 +446,20 @@ static void skb_recv_done(struct virtque
}
}
+static void virtnet_napi_enable(struct virtnet_info *vi)
+{
+ napi_enable(&vi->napi);
+
+ /* If all buffers were filled by other side before we napi_enabled, we
+ * won't get another interrupt, so process any outstanding packets
+ * now. virtnet_poll wants re-enable the queue, so we disable here.
+ * We synchronize against interrupts via NAPI_STATE_SCHED */
+ if (napi_schedule_prep(&vi->napi)) {
+ virtqueue_disable_cb(vi->rvq);
+ __napi_schedule(&vi->napi);
+ }
+}
+
static void refill_work(struct work_struct *work)
{
struct virtnet_info *vi;
@@ -454,7 +468,7 @@ static void refill_work(struct work_stru
vi = container_of(work, struct virtnet_info, refill.work);
napi_disable(&vi->napi);
still_empty = !try_fill_recv(vi, GFP_KERNEL);
- napi_enable(&vi->napi);
+ virtnet_napi_enable(vi);
/* In theory, this can happen: if we don't get any buffers in
* we will *never* try to fill again. */
@@ -638,16 +652,7 @@ static int virtnet_open(struct net_devic
{
struct virtnet_info *vi = netdev_priv(dev);
- napi_enable(&vi->napi);
-
- /* If all buffers were filled by other side before we napi_enabled, we
- * won't get another interrupt, so process any outstanding packets
- * now. virtnet_poll wants re-enable the queue, so we disable here.
- * We synchronize against interrupts via NAPI_STATE_SCHED */
- if (napi_schedule_prep(&vi->napi)) {
- virtqueue_disable_cb(vi->rvq);
- __napi_schedule(&vi->napi);
- }
+ virtnet_napi_enable(vi);
return 0;
}
^ permalink raw reply
* Re: [PATCH] virtio-net: add schedule check to napi_enable call in refill_work
From: Bruce Rogers @ 2011-02-10 2:44 UTC (permalink / raw)
To: virtualization, Rusty Russell; +Cc: netdev
In-Reply-To: <201102101201.19656.rusty@rustcorp.com.au>
>>> On 2/9/2011 at 06:31 PM, Rusty Russell <rusty@rustcorp.com.au> wrote:
> On Thu, 10 Feb 2011 06:59:25 am Ken Stailey wrote:
>> Justification:
>>
>> Impact: Under heavy network I/O load virtio-net driver crashes making VM
> guest unusable.
>
> Hmm, this went badly wrong. I acked this patch, and it was mailed to
> netdev six months ago.
>
> Bruce's patch used spaces instead of tabs, but that should not have caused
> it to be dropped. I've taken that and ported it forwards, will repost now.
>
> Thanks for picking this up off the floor!
> Rusty.
Thanks for taking care of that!
Bruce
^ permalink raw reply
* Re: [PATCH] virtio_net: Add schedule check to napi_enable call
From: Michael S. Tsirkin @ 2011-02-10 17:57 UTC (permalink / raw)
To: Rusty Russell; +Cc: Herbert Xu, virtualization, David Miller, netdev
In-Reply-To: <201102101232.50526.rusty@rustcorp.com.au>
On Thu, Feb 10, 2011 at 12:32:50PM +1030, Rusty Russell wrote:
> From: "Bruce Rogers" <brogers@novell.com>
>
> Under harsh testing conditions, including low memory, the guest would
> stop receiving packets. With this patch applied we no longer see any
> problems in the driver while performing these tests for extended periods
> of time.
>
> Make sure napi is scheduled subsequent to each napi_enable.
>
> Signed-off-by: Bruce Rogers <brogers@novell.com>
> Signed-off-by: Olaf Kirch <okir@suse.de>
> Cc: stable@kernel.org
> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Rusty, so this is 2.6.38 material - you'll send this to Linus? Or DaveM?
> ---
> drivers/net/virtio_net.c | 27 ++++++++++++++++-----------
> 1 file changed, 16 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -446,6 +446,20 @@ static void skb_recv_done(struct virtque
> }
> }
>
> +static void virtnet_napi_enable(struct virtnet_info *vi)
> +{
> + napi_enable(&vi->napi);
> +
> + /* If all buffers were filled by other side before we napi_enabled, we
> + * won't get another interrupt, so process any outstanding packets
> + * now. virtnet_poll wants re-enable the queue, so we disable here.
> + * We synchronize against interrupts via NAPI_STATE_SCHED */
> + if (napi_schedule_prep(&vi->napi)) {
> + virtqueue_disable_cb(vi->rvq);
> + __napi_schedule(&vi->napi);
> + }
> +}
> +
> static void refill_work(struct work_struct *work)
> {
> struct virtnet_info *vi;
> @@ -454,7 +468,7 @@ static void refill_work(struct work_stru
> vi = container_of(work, struct virtnet_info, refill.work);
> napi_disable(&vi->napi);
> still_empty = !try_fill_recv(vi, GFP_KERNEL);
> - napi_enable(&vi->napi);
> + virtnet_napi_enable(vi);
>
> /* In theory, this can happen: if we don't get any buffers in
> * we will *never* try to fill again. */
> @@ -638,16 +652,7 @@ static int virtnet_open(struct net_devic
> {
> struct virtnet_info *vi = netdev_priv(dev);
>
> - napi_enable(&vi->napi);
> -
> - /* If all buffers were filled by other side before we napi_enabled, we
> - * won't get another interrupt, so process any outstanding packets
> - * now. virtnet_poll wants re-enable the queue, so we disable here.
> - * We synchronize against interrupts via NAPI_STATE_SCHED */
> - if (napi_schedule_prep(&vi->napi)) {
> - virtqueue_disable_cb(vi->rvq);
> - __napi_schedule(&vi->napi);
> - }
> + virtnet_napi_enable(vi);
> return 0;
> }
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH] virtio_net: Add schedule check to napi_enable call
From: David Miller @ 2011-02-10 19:01 UTC (permalink / raw)
To: mst; +Cc: herbert, virtualization, netdev
In-Reply-To: <20110210175725.GA9674@redhat.com>
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Thu, 10 Feb 2011 19:57:26 +0200
> On Thu, Feb 10, 2011 at 12:32:50PM +1030, Rusty Russell wrote:
>> From: "Bruce Rogers" <brogers@novell.com>
>>
>> Under harsh testing conditions, including low memory, the guest would
>> stop receiving packets. With this patch applied we no longer see any
>> problems in the driver while performing these tests for extended periods
>> of time.
>>
>> Make sure napi is scheduled subsequent to each napi_enable.
>>
>> Signed-off-by: Bruce Rogers <brogers@novell.com>
>> Signed-off-by: Olaf Kirch <okir@suse.de>
>> Cc: stable@kernel.org
>> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
>
> Rusty, so this is 2.6.38 material - you'll send this to Linus? Or DaveM?
Don't worry I'll apply this to net-2.6, thanks.
^ permalink raw reply
* Re: [PATCH] virtio_net: Add schedule check to napi_enable call
From: David Miller @ 2011-02-10 19:03 UTC (permalink / raw)
To: rusty; +Cc: herbert, virtualization, netdev
In-Reply-To: <201102101232.50526.rusty@rustcorp.com.au>
From: Rusty Russell <rusty@rustcorp.com.au>
Date: Thu, 10 Feb 2011 12:32:50 +1030
> From: "Bruce Rogers" <brogers@novell.com>
>
> Under harsh testing conditions, including low memory, the guest would
> stop receiving packets. With this patch applied we no longer see any
> problems in the driver while performing these tests for extended periods
> of time.
>
> Make sure napi is scheduled subsequent to each napi_enable.
>
> Signed-off-by: Bruce Rogers <brogers@novell.com>
> Signed-off-by: Olaf Kirch <okir@suse.de>
> Cc: stable@kernel.org
> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Applied, thanks everyone.
^ permalink raw reply
* [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: K. Y. Srinivasan @ 2011-02-11 17:59 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, virtualization; +Cc: K. Y. Srinivasan
In preperation for getting rid of the osd.[ch] files;
change all page allocation/free functions to use native interfaces.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Hank Janssen <hjanssen@microsoft.com>
---
drivers/staging/hv/channel.c | 12 +++++++-----
drivers/staging/hv/connection.c | 13 ++++++++-----
drivers/staging/hv/hv.c | 15 ++++++++++-----
drivers/staging/hv/netvsc.c | 14 ++++++++------
4 files changed, 33 insertions(+), 21 deletions(-)
diff --git a/drivers/staging/hv/channel.c b/drivers/staging/hv/channel.c
index ba9afda..6c292e6 100644
--- a/drivers/staging/hv/channel.c
+++ b/drivers/staging/hv/channel.c
@@ -180,8 +180,9 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
newchannel->channel_callback_context = context;
/* Allocate the ring buffer */
- out = osd_page_alloc((send_ringbuffer_size + recv_ringbuffer_size)
- >> PAGE_SHIFT);
+ out = (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO,
+ get_order(send_ringbuffer_size + recv_ringbuffer_size));
+
if (!out)
return -ENOMEM;
@@ -300,8 +301,8 @@ Cleanup:
errorout:
ringbuffer_cleanup(&newchannel->outbound);
ringbuffer_cleanup(&newchannel->inbound);
- osd_page_free(out, (send_ringbuffer_size + recv_ringbuffer_size)
- >> PAGE_SHIFT);
+ free_pages((unsigned long)out,
+ get_order(send_ringbuffer_size + recv_ringbuffer_size));
kfree(openInfo);
return err;
}
@@ -686,7 +687,8 @@ void vmbus_close(struct vmbus_channel *channel)
ringbuffer_cleanup(&channel->outbound);
ringbuffer_cleanup(&channel->inbound);
- osd_page_free(channel->ringbuffer_pages, channel->ringbuffer_pagecount);
+ free_pages((unsigned long)channel->ringbuffer_pages,
+ get_order(channel->ringbuffer_pagecount * PAGE_SIZE));
kfree(info);
diff --git a/drivers/staging/hv/connection.c b/drivers/staging/hv/connection.c
index b3ac66e..ed0976a 100644
--- a/drivers/staging/hv/connection.c
+++ b/drivers/staging/hv/connection.c
@@ -66,7 +66,8 @@ int vmbus_connect(void)
* Setup the vmbus event connection for channel interrupt
* abstraction stuff
*/
- vmbus_connection.int_page = osd_page_alloc(1);
+ vmbus_connection.int_page =
+ (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO, 0);
if (vmbus_connection.int_page == NULL) {
ret = -1;
goto Cleanup;
@@ -81,7 +82,8 @@ int vmbus_connect(void)
* Setup the monitor notification facility. The 1st page for
* parent->child and the 2nd page for child->parent
*/
- vmbus_connection.monitor_pages = osd_page_alloc(2);
+ vmbus_connection.monitor_pages =
+ (void *)__get_free_pages((GFP_KERNEL|__GFP_ZERO), 1);
if (vmbus_connection.monitor_pages == NULL) {
ret = -1;
goto Cleanup;
@@ -162,12 +164,12 @@ Cleanup:
destroy_workqueue(vmbus_connection.work_queue);
if (vmbus_connection.int_page) {
- osd_page_free(vmbus_connection.int_page, 1);
+ free_pages((unsigned long)vmbus_connection.int_page, 0);
vmbus_connection.int_page = NULL;
}
if (vmbus_connection.monitor_pages) {
- osd_page_free(vmbus_connection.monitor_pages, 2);
+ free_pages((unsigned long)vmbus_connection.monitor_pages, 1);
vmbus_connection.monitor_pages = NULL;
}
@@ -203,7 +205,8 @@ int vmbus_disconnect(void)
if (ret != 0)
goto Cleanup;
- osd_page_free(vmbus_connection.int_page, 1);
+ free_pages((unsigned long)vmbus_connection.int_page, 0);
+ free_pages((unsigned long)vmbus_connection.monitor_pages, 1);
/* TODO: iterate thru the msg list and free up */
destroy_workqueue(vmbus_connection.work_queue);
diff --git a/drivers/staging/hv/hv.c b/drivers/staging/hv/hv.c
index 021acba..419b4d6 100644
--- a/drivers/staging/hv/hv.c
+++ b/drivers/staging/hv/hv.c
@@ -230,7 +230,12 @@ int hv_init(void)
* Allocate the hypercall page memory
* virtaddr = osd_page_alloc(1);
*/
- virtaddr = osd_virtual_alloc_exec(PAGE_SIZE);
+#ifdef __x86_64__
+ virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC);
+#else
+ virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
+ __pgprot(__PAGE_KERNEL & (~_PAGE_NX)));
+#endif
if (!virtaddr) {
DPRINT_ERR(VMBUS,
@@ -462,10 +467,10 @@ void hv_synic_init(void *irqarg)
Cleanup:
if (hv_context.synic_event_page[cpu])
- osd_page_free(hv_context.synic_event_page[cpu], 1);
+ free_page((unsigned long)hv_context.synic_event_page[cpu]);
if (hv_context.synic_message_page[cpu])
- osd_page_free(hv_context.synic_message_page[cpu], 1);
+ free_page((unsigned long)hv_context.synic_message_page[cpu]);
return;
}
@@ -502,6 +507,6 @@ void hv_synic_cleanup(void *arg)
wrmsrl(HV_X64_MSR_SIEFP, siefp.as_uint64);
- osd_page_free(hv_context.synic_message_page[cpu], 1);
- osd_page_free(hv_context.synic_event_page[cpu], 1);
+ free_page((unsigned long)hv_context.synic_message_page[cpu]);
+ free_page((unsigned long)hv_context.synic_event_page[cpu]);
}
diff --git a/drivers/staging/hv/netvsc.c b/drivers/staging/hv/netvsc.c
index 4319363..a271aa7 100644
--- a/drivers/staging/hv/netvsc.c
+++ b/drivers/staging/hv/netvsc.c
@@ -223,7 +223,8 @@ static int netvsc_init_recv_buf(struct hv_device *device)
/* ASSERT((netDevice->ReceiveBufferSize & (PAGE_SIZE - 1)) == 0); */
net_device->recv_buf =
- osd_page_alloc(net_device->recv_buf_size >> PAGE_SHIFT);
+ (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO,
+ get_order(net_device->recv_buf_size));
if (!net_device->recv_buf) {
DPRINT_ERR(NETVSC,
"unable to allocate receive buffer of size %d",
@@ -360,7 +361,8 @@ static int netvsc_init_send_buf(struct hv_device *device)
/* ASSERT((netDevice->SendBufferSize & (PAGE_SIZE - 1)) == 0); */
net_device->send_buf =
- osd_page_alloc(net_device->send_buf_size >> PAGE_SHIFT);
+ (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO,
+ get_order(net_device->send_buf_size));
if (!net_device->send_buf) {
DPRINT_ERR(NETVSC, "unable to allocate send buffer of size %d",
net_device->send_buf_size);
@@ -498,8 +500,8 @@ static int netvsc_destroy_recv_buf(struct netvsc_device *net_device)
DPRINT_INFO(NETVSC, "Freeing up receive buffer...");
/* Free up the receive buffer */
- osd_page_free(net_device->recv_buf,
- net_device->recv_buf_size >> PAGE_SHIFT);
+ free_pages((unsigned long)net_device->recv_buf,
+ get_order(net_device->recv_buf_size));
net_device->recv_buf = NULL;
}
@@ -574,8 +576,8 @@ static int netvsc_destroy_send_buf(struct netvsc_device *net_device)
DPRINT_INFO(NETVSC, "Freeing up send buffer...");
/* Free up the receive buffer */
- osd_page_free(net_device->send_buf,
- net_device->send_buf_size >> PAGE_SHIFT);
+ free_pages((unsigned long)net_device->send_buf,
+ get_order(net_device->send_buf_size));
net_device->send_buf = NULL;
}
--
1.5.5.6
^ permalink raw reply related
* [PATCH 2/3]: Staging: hv: Use native wait primitives
From: K. Y. Srinivasan @ 2011-02-11 17:59 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, virtualization; +Cc: K. Y. Srinivasan
In preperation for getting rid of the osd layer; change
the code to use native wait interfaces. As part of this,
fixed the buggy implementation in the osd_wait_primitive
where the condition was cleared potentially after the
condition was signalled.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Hank Janssen <hjanssen@microsoft.com>
---
drivers/staging/hv/channel.c | 58 +++++++++-----------
drivers/staging/hv/channel_mgmt.c | 46 +++++++---------
drivers/staging/hv/channel_mgmt.h | 4 +-
drivers/staging/hv/connection.c | 28 +++++++---
drivers/staging/hv/netvsc.c | 102 ++++++++++++++++++-----------------
drivers/staging/hv/netvsc.h | 3 +-
drivers/staging/hv/rndis_filter.c | 38 +++++++++----
drivers/staging/hv/storvsc.c | 98 ++++++++++++++++++++--------------
drivers/staging/hv/vmbus_private.h | 3 +-
9 files changed, 208 insertions(+), 172 deletions(-)
diff --git a/drivers/staging/hv/channel.c b/drivers/staging/hv/channel.c
index 6c292e6..5a0923c 100644
--- a/drivers/staging/hv/channel.c
+++ b/drivers/staging/hv/channel.c
@@ -19,6 +19,8 @@
* Hank Janssen <hjanssen@microsoft.com>
*/
#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/module.h>
@@ -243,11 +245,7 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
goto errorout;
}
- openInfo->waitevent = osd_waitevent_create();
- if (!openInfo->waitevent) {
- err = -ENOMEM;
- goto errorout;
- }
+ init_waitqueue_head(&openInfo->waitevent);
openMsg = (struct vmbus_channel_open_channel *)openInfo->msg;
openMsg->header.msgtype = CHANNELMSG_OPENCHANNEL;
@@ -280,8 +278,15 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
goto Cleanup;
}
- /* FIXME: Need to time-out here */
- osd_waitevent_wait(openInfo->waitevent);
+ openInfo->wait_condition = 0;
+ wait_event_timeout(openInfo->waitevent,
+ openInfo->wait_condition,
+ msecs_to_jiffies(1000));
+ if (openInfo->wait_condition == 0) {
+ err = -ETIMEDOUT;
+ goto errorout;
+ }
+
if (openInfo->response.open_result.status == 0)
DPRINT_INFO(VMBUS, "channel <%p> open success!!", newchannel);
@@ -294,7 +299,6 @@ Cleanup:
list_del(&openInfo->msglistentry);
spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
- kfree(openInfo->waitevent);
kfree(openInfo);
return 0;
@@ -509,11 +513,7 @@ int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer,
if (ret)
return ret;
- msginfo->waitevent = osd_waitevent_create();
- if (!msginfo->waitevent) {
- ret = -ENOMEM;
- goto Cleanup;
- }
+ init_waitqueue_head(&msginfo->waitevent);
gpadlmsg = (struct vmbus_channel_gpadl_header *)msginfo->msg;
gpadlmsg->header.msgtype = CHANNELMSG_GPADL_HEADER;
@@ -533,6 +533,7 @@ int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer,
DPRINT_DBG(VMBUS, "Sending GPADL Header - len %zd",
msginfo->msgsize - sizeof(*msginfo));
+ msginfo->wait_condition = 0;
ret = vmbus_post_msg(gpadlmsg, msginfo->msgsize -
sizeof(*msginfo));
if (ret != 0) {
@@ -566,7 +567,11 @@ int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer,
}
}
- osd_waitevent_wait(msginfo->waitevent);
+ wait_event_timeout(msginfo->waitevent,
+ msginfo->wait_condition,
+ msecs_to_jiffies(1000));
+ BUG_ON(msginfo->wait_condition == 0);
+
/* At this point, we received the gpadl created msg */
DPRINT_DBG(VMBUS, "Received GPADL created "
@@ -582,7 +587,6 @@ Cleanup:
list_del(&msginfo->msglistentry);
spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
- kfree(msginfo->waitevent);
kfree(msginfo);
return ret;
}
@@ -605,11 +609,7 @@ int vmbus_teardown_gpadl(struct vmbus_channel *channel, u32 gpadl_handle)
if (!info)
return -ENOMEM;
- info->waitevent = osd_waitevent_create();
- if (!info->waitevent) {
- kfree(info);
- return -ENOMEM;
- }
+ init_waitqueue_head(&info->waitevent);
msg = (struct vmbus_channel_gpadl_teardown *)info->msg;
@@ -621,22 +621,20 @@ int vmbus_teardown_gpadl(struct vmbus_channel *channel, u32 gpadl_handle)
list_add_tail(&info->msglistentry,
&vmbus_connection.chn_msg_list);
spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
-
+ info->wait_condition = 0;
ret = vmbus_post_msg(msg,
sizeof(struct vmbus_channel_gpadl_teardown));
- if (ret != 0) {
- /* TODO: */
- /* something... */
- }
- osd_waitevent_wait(info->waitevent);
+ BUG_ON(ret != 0);
+ wait_event_timeout(info->waitevent,
+ info->wait_condition, msecs_to_jiffies(1000));
+ BUG_ON(info->wait_condition == 0);
/* Received a torndown response */
spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags);
list_del(&info->msglistentry);
spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
- kfree(info->waitevent);
kfree(info);
return ret;
}
@@ -664,18 +662,14 @@ void vmbus_close(struct vmbus_channel *channel)
if (!info)
return;
- /* info->waitEvent = osd_waitevent_create(); */
msg = (struct vmbus_channel_close_channel *)info->msg;
msg->header.msgtype = CHANNELMSG_CLOSECHANNEL;
msg->child_relid = channel->offermsg.child_relid;
ret = vmbus_post_msg(msg, sizeof(struct vmbus_channel_close_channel));
- if (ret != 0) {
- /* TODO: */
- /* something... */
- }
+ BUG_ON(ret != 0);
/* Tear down the gpadl for the channel's ring buffer */
if (channel->ringbuffer_gpadlhandle)
vmbus_teardown_gpadl(channel,
diff --git a/drivers/staging/hv/channel_mgmt.c b/drivers/staging/hv/channel_mgmt.c
index a9c9d49..da1e56a 100644
--- a/drivers/staging/hv/channel_mgmt.c
+++ b/drivers/staging/hv/channel_mgmt.c
@@ -19,6 +19,8 @@
* Hank Janssen <hjanssen@microsoft.com>
*/
#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/list.h>
@@ -593,7 +595,8 @@ static void vmbus_onopen_result(struct vmbus_channel_message_header *hdr)
memcpy(&msginfo->response.open_result,
result,
sizeof(struct vmbus_channel_open_result));
- osd_waitevent_set(msginfo->waitevent);
+ msginfo->wait_condition = 1;
+ wake_up(&msginfo->waitevent);
break;
}
}
@@ -643,7 +646,8 @@ static void vmbus_ongpadl_created(struct vmbus_channel_message_header *hdr)
memcpy(&msginfo->response.gpadl_created,
gpadlcreated,
sizeof(struct vmbus_channel_gpadl_created));
- osd_waitevent_set(msginfo->waitevent);
+ msginfo->wait_condition = 1;
+ wake_up(&msginfo->waitevent);
break;
}
}
@@ -689,7 +693,8 @@ static void vmbus_ongpadl_torndown(
memcpy(&msginfo->response.gpadl_torndown,
gpadl_torndown,
sizeof(struct vmbus_channel_gpadl_torndown));
- osd_waitevent_set(msginfo->waitevent);
+ msginfo->wait_condition = 1;
+ wake_up(&msginfo->waitevent);
break;
}
}
@@ -730,7 +735,8 @@ static void vmbus_onversion_response(
memcpy(&msginfo->response.version_response,
version_response,
sizeof(struct vmbus_channel_version_response));
- osd_waitevent_set(msginfo->waitevent);
+ msginfo->wait_condition = 1;
+ wake_up(&msginfo->waitevent);
}
}
spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
@@ -805,44 +811,34 @@ int vmbus_request_offers(void)
if (!msginfo)
return -ENOMEM;
- msginfo->waitevent = osd_waitevent_create();
- if (!msginfo->waitevent) {
- kfree(msginfo);
- return -ENOMEM;
- }
+ init_waitqueue_head(&msginfo->waitevent);
msg = (struct vmbus_channel_message_header *)msginfo->msg;
msg->msgtype = CHANNELMSG_REQUESTOFFERS;
- /*SpinlockAcquire(gVmbusConnection.channelMsgLock);
- INSERT_TAIL_LIST(&gVmbusConnection.channelMsgList,
- &msgInfo->msgListEntry);
- SpinlockRelease(gVmbusConnection.channelMsgLock);*/
ret = vmbus_post_msg(msg,
sizeof(struct vmbus_channel_message_header));
if (ret != 0) {
DPRINT_ERR(VMBUS, "Unable to request offers - %d", ret);
- /*SpinlockAcquire(gVmbusConnection.channelMsgLock);
- REMOVE_ENTRY_LIST(&msgInfo->msgListEntry);
- SpinlockRelease(gVmbusConnection.channelMsgLock);*/
+ goto cleanup;
+ }
- goto Cleanup;
+ msginfo->wait_condition = 0;
+ wait_event_timeout(msginfo->waitevent, msginfo->wait_condition,
+ msecs_to_jiffies(1000));
+ if (msginfo->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
}
- /* osd_waitevent_wait(msgInfo->waitEvent); */
- /*SpinlockAcquire(gVmbusConnection.channelMsgLock);
- REMOVE_ENTRY_LIST(&msgInfo->msgListEntry);
- SpinlockRelease(gVmbusConnection.channelMsgLock);*/
-Cleanup:
- if (msginfo) {
- kfree(msginfo->waitevent);
+cleanup:
+ if (msginfo)
kfree(msginfo);
- }
return ret;
}
diff --git a/drivers/staging/hv/channel_mgmt.h b/drivers/staging/hv/channel_mgmt.h
index fe40bf2..3368bf1 100644
--- a/drivers/staging/hv/channel_mgmt.h
+++ b/drivers/staging/hv/channel_mgmt.h
@@ -289,8 +289,8 @@ struct vmbus_channel_msginfo {
struct list_head submsglist;
/* Synchronize the request/response if needed */
- struct osd_waitevent *waitevent;
-
+ int wait_condition;
+ wait_queue_head_t waitevent;
union {
struct vmbus_channel_version_supported version_supported;
struct vmbus_channel_open_result open_result;
diff --git a/drivers/staging/hv/connection.c b/drivers/staging/hv/connection.c
index ed0976a..51dd362 100644
--- a/drivers/staging/hv/connection.c
+++ b/drivers/staging/hv/connection.c
@@ -21,6 +21,8 @@
*
*/
#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>
@@ -97,11 +99,7 @@ int vmbus_connect(void)
goto Cleanup;
}
- msginfo->waitevent = osd_waitevent_create();
- if (!msginfo->waitevent) {
- ret = -ENOMEM;
- goto Cleanup;
- }
+ init_waitqueue_head(&msginfo->waitevent);
msg = (struct vmbus_channel_initiate_contact *)msginfo->msg;
@@ -131,14 +129,30 @@ int vmbus_connect(void)
ret = vmbus_post_msg(msg,
sizeof(struct vmbus_channel_initiate_contact));
if (ret != 0) {
+ spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags);
list_del(&msginfo->msglistentry);
+ spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock,
+ flags);
goto Cleanup;
}
/* Wait for the connection response */
- osd_waitevent_wait(msginfo->waitevent);
+ msginfo->wait_condition = 0;
+ wait_event_timeout(msginfo->waitevent, msginfo->wait_condition,
+ msecs_to_jiffies(1000));
+ if (msginfo->wait_condition == 0) {
+ spin_lock_irqsave(&vmbus_connection.channelmsg_lock,
+ flags);
+ list_del(&msginfo->msglistentry);
+ spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock,
+ flags);
+ ret = -ETIMEDOUT;
+ goto Cleanup;
+ }
+ spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags);
list_del(&msginfo->msglistentry);
+ spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
/* Check if successful */
if (msginfo->response.version_response.version_supported) {
@@ -153,7 +167,6 @@ int vmbus_connect(void)
goto Cleanup;
}
- kfree(msginfo->waitevent);
kfree(msginfo);
return 0;
@@ -174,7 +187,6 @@ Cleanup:
}
if (msginfo) {
- kfree(msginfo->waitevent);
kfree(msginfo);
}
diff --git a/drivers/staging/hv/netvsc.c b/drivers/staging/hv/netvsc.c
index a271aa7..7233564 100644
--- a/drivers/staging/hv/netvsc.c
+++ b/drivers/staging/hv/netvsc.c
@@ -19,6 +19,8 @@
* Hank Janssen <hjanssen@microsoft.com>
*/
#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
#include <linux/mm.h>
#include <linux/delay.h>
#include <linux/io.h>
@@ -230,7 +232,7 @@ static int netvsc_init_recv_buf(struct hv_device *device)
"unable to allocate receive buffer of size %d",
net_device->recv_buf_size);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
/* page-aligned buffer */
/* ASSERT(((unsigned long)netDevice->ReceiveBuffer & (PAGE_SIZE - 1)) == */
@@ -249,10 +251,9 @@ static int netvsc_init_recv_buf(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(NETVSC,
"unable to establish receive buffer's gpadl");
- goto Cleanup;
+ goto cleanup;
}
- /* osd_waitevent_wait(ext->ChannelInitEvent); */
/* Notify the NetVsp of the gpadl handle */
DPRINT_INFO(NETVSC, "Sending NvspMessage1TypeSendReceiveBuffer...");
@@ -268,6 +269,7 @@ static int netvsc_init_recv_buf(struct hv_device *device)
send_recv_buf.id = NETVSC_RECEIVE_BUFFER_ID;
/* Send the gpadl notification request */
+ net_device->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, init_packet,
sizeof(struct nvsp_message),
(unsigned long)init_packet,
@@ -276,10 +278,14 @@ static int netvsc_init_recv_buf(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(NETVSC,
"unable to send receive buffer's gpadl to netvsp");
- goto Cleanup;
+ goto cleanup;
}
- osd_waitevent_wait(net_device->channel_init_event);
+ wait_event_timeout(net_device->channel_init_wait,
+ net_device->wait_condition,
+ msecs_to_jiffies(1000));
+ BUG_ON(net_device->wait_condition == 0);
+
/* Check the response */
if (init_packet->msg.v1_msg.
@@ -289,7 +295,7 @@ static int netvsc_init_recv_buf(struct hv_device *device)
init_packet->msg.v1_msg.
send_recv_buf_complete.status);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
/* Parse the response */
@@ -303,7 +309,7 @@ static int netvsc_init_recv_buf(struct hv_device *device)
* sizeof(struct nvsp_1_receive_buffer_section), GFP_KERNEL);
if (net_device->recv_section == NULL) {
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
memcpy(net_device->recv_section,
@@ -327,15 +333,15 @@ static int netvsc_init_recv_buf(struct hv_device *device)
if (net_device->recv_section_cnt != 1 ||
net_device->recv_section->offset != 0) {
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
- goto Exit;
+ goto exit;
-Cleanup:
+cleanup:
netvsc_destroy_recv_buf(net_device);
-Exit:
+exit:
put_net_device(device);
return ret;
}
@@ -354,7 +360,7 @@ static int netvsc_init_send_buf(struct hv_device *device)
}
if (net_device->send_buf_size <= 0) {
ret = -EINVAL;
- goto Cleanup;
+ goto cleanup;
}
/* page-size grandularity */
@@ -367,7 +373,7 @@ static int netvsc_init_send_buf(struct hv_device *device)
DPRINT_ERR(NETVSC, "unable to allocate send buffer of size %d",
net_device->send_buf_size);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
/* page-aligned buffer */
/* ASSERT(((unsigned long)netDevice->SendBuffer & (PAGE_SIZE - 1)) == 0); */
@@ -384,11 +390,9 @@ static int netvsc_init_send_buf(struct hv_device *device)
&net_device->send_buf_gpadl_handle);
if (ret != 0) {
DPRINT_ERR(NETVSC, "unable to establish send buffer's gpadl");
- goto Cleanup;
+ goto cleanup;
}
- /* osd_waitevent_wait(ext->ChannelInitEvent); */
-
/* Notify the NetVsp of the gpadl handle */
DPRINT_INFO(NETVSC, "Sending NvspMessage1TypeSendSendBuffer...");
@@ -403,6 +407,7 @@ static int netvsc_init_send_buf(struct hv_device *device)
NETVSC_SEND_BUFFER_ID;
/* Send the gpadl notification request */
+ net_device->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, init_packet,
sizeof(struct nvsp_message),
(unsigned long)init_packet,
@@ -411,10 +416,13 @@ static int netvsc_init_send_buf(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(NETVSC,
"unable to send receive buffer's gpadl to netvsp");
- goto Cleanup;
+ goto cleanup;
}
- osd_waitevent_wait(net_device->channel_init_event);
+ wait_event_timeout(net_device->channel_init_wait,
+ net_device->wait_condition,
+ msecs_to_jiffies(1000));
+ BUG_ON(net_device->wait_condition == 0);
/* Check the response */
if (init_packet->msg.v1_msg.
@@ -424,18 +432,18 @@ static int netvsc_init_send_buf(struct hv_device *device)
init_packet->msg.v1_msg.
send_send_buf_complete.status);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
net_device->send_section_size = init_packet->
msg.v1_msg.send_send_buf_complete.section_size;
- goto Exit;
+ goto exit;
-Cleanup:
+cleanup:
netvsc_destroy_send_buf(net_device);
-Exit:
+exit:
put_net_device(device);
return ret;
}
@@ -611,6 +619,7 @@ static int netvsc_connect_vsp(struct hv_device *device)
DPRINT_INFO(NETVSC, "Sending NvspMessageTypeInit...");
/* Send the init request */
+ net_device->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, init_packet,
sizeof(struct nvsp_message),
(unsigned long)init_packet,
@@ -619,10 +628,16 @@ static int netvsc_connect_vsp(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(NETVSC, "unable to send NvspMessageTypeInit");
- goto Cleanup;
+ goto cleanup;
}
- osd_waitevent_wait(net_device->channel_init_event);
+ wait_event_timeout(net_device->channel_init_wait,
+ net_device->wait_condition,
+ msecs_to_jiffies(1000));
+ if (net_device->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
/* Now, check the response */
/* ASSERT(initPacket->Messages.InitMessages.InitComplete.MaximumMdlChainLength <= MAX_MULTIPAGE_BUFFER_COUNT); */
@@ -637,7 +652,7 @@ static int netvsc_connect_vsp(struct hv_device *device)
"unable to initialize with netvsp (status 0x%x)",
init_packet->msg.init_msg.init_complete.status);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
if (init_packet->msg.init_msg.init_complete.
@@ -647,7 +662,7 @@ static int netvsc_connect_vsp(struct hv_device *device)
init_packet->msg.init_msg.
init_complete.negotiated_protocol_ver);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
DPRINT_INFO(NETVSC, "Sending NvspMessage1TypeSendNdisVersion...");
@@ -666,29 +681,22 @@ static int netvsc_connect_vsp(struct hv_device *device)
/* Send the init request */
ret = vmbus_sendpacket(device->channel, init_packet,
- sizeof(struct nvsp_message),
- (unsigned long)init_packet,
- VM_PKT_DATA_INBAND, 0);
+ sizeof(struct nvsp_message),
+ (unsigned long)init_packet,
+ VM_PKT_DATA_INBAND, 0);
if (ret != 0) {
DPRINT_ERR(NETVSC,
"unable to send NvspMessage1TypeSendNdisVersion");
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
- /*
- * BUGBUG - We have to wait for the above msg since the
- * netvsp uses KMCL which acknowledges packet (completion
- * packet) since our Vmbus always set the
- * VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED flag
- */
- /* osd_waitevent_wait(NetVscChannel->ChannelInitEvent); */
/* Post the big receive buffer to NetVSP */
ret = netvsc_init_recv_buf(device);
if (ret == 0)
ret = netvsc_init_send_buf(device);
-Cleanup:
+cleanup:
put_net_device(device);
return ret;
}
@@ -715,7 +723,7 @@ static int netvsc_device_add(struct hv_device *device, void *additional_info)
net_device = alloc_net_device(device);
if (!net_device) {
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
DPRINT_DBG(NETVSC, "netvsc channel object allocated - %p", net_device);
@@ -741,11 +749,7 @@ static int netvsc_device_add(struct hv_device *device, void *additional_info)
list_add_tail(&packet->list_ent,
&net_device->recv_pkt_list);
}
- net_device->channel_init_event = osd_waitevent_create();
- if (!net_device->channel_init_event) {
- ret = -ENOMEM;
- goto Cleanup;
- }
+ init_waitqueue_head(&net_device->channel_init_wait);
/* Open the channel */
ret = vmbus_open(device->channel, net_driver->ring_buf_size,
@@ -755,7 +759,7 @@ static int netvsc_device_add(struct hv_device *device, void *additional_info)
if (ret != 0) {
DPRINT_ERR(NETVSC, "unable to open channel: %d", ret);
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
/* Channel is opened */
@@ -778,11 +782,9 @@ close:
/* Now, we can close the channel safely */
vmbus_close(device->channel);
-Cleanup:
+cleanup:
if (net_device) {
- kfree(net_device->channel_init_event);
-
list_for_each_entry_safe(packet, pos,
&net_device->recv_pkt_list,
list_ent) {
@@ -847,7 +849,6 @@ static int netvsc_device_remove(struct hv_device *device)
kfree(netvsc_packet);
}
- kfree(net_device->channel_init_event);
free_net_device(net_device);
return 0;
}
@@ -887,7 +888,8 @@ static void netvsc_send_completion(struct hv_device *device,
/* Copy the response back */
memcpy(&net_device->channel_init_pkt, nvsp_packet,
sizeof(struct nvsp_message));
- osd_waitevent_set(net_device->channel_init_event);
+ net_device->wait_condition = 1;
+ wake_up(&net_device->channel_init_wait);
} else if (nvsp_packet->hdr.msg_type ==
NVSP_MSG1_TYPE_SEND_RNDIS_PKT_COMPLETE) {
/* Get the send context */
diff --git a/drivers/staging/hv/netvsc.h b/drivers/staging/hv/netvsc.h
index 5f6dcf1..45d24b9 100644
--- a/drivers/staging/hv/netvsc.h
+++ b/drivers/staging/hv/netvsc.h
@@ -318,7 +318,8 @@ struct netvsc_device {
struct nvsp_1_receive_buffer_section *recv_section;
/* Used for NetVSP initialization protocol */
- struct osd_waitevent *channel_init_event;
+ int wait_condition;
+ wait_queue_head_t channel_init_wait;
struct nvsp_message channel_init_pkt;
struct nvsp_message revoke_packet;
diff --git a/drivers/staging/hv/rndis_filter.c b/drivers/staging/hv/rndis_filter.c
index 287e12e..e3bf004 100644
--- a/drivers/staging/hv/rndis_filter.c
+++ b/drivers/staging/hv/rndis_filter.c
@@ -19,6 +19,8 @@
* Hank Janssen <hjanssen@microsoft.com>
*/
#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
#include <linux/highmem.h>
#include <linux/slab.h>
#include <linux/io.h>
@@ -57,7 +59,8 @@ struct rndis_device {
struct rndis_request {
struct list_head list_ent;
- struct osd_waitevent *waitevent;
+ int wait_condition;
+ wait_queue_head_t wait_event;
/*
* FIXME: We assumed a fixed size response here. If we do ever need to
@@ -129,11 +132,7 @@ static struct rndis_request *get_rndis_request(struct rndis_device *dev,
if (!request)
return NULL;
- request->waitevent = osd_waitevent_create();
- if (!request->waitevent) {
- kfree(request);
- return NULL;
- }
+ init_waitqueue_head(&request->wait_event);
rndis_msg = &request->request_msg;
rndis_msg->ndis_msg_type = msg_type;
@@ -164,7 +163,6 @@ static void put_rndis_request(struct rndis_device *dev,
list_del(&req->list_ent);
spin_unlock_irqrestore(&dev->request_lock, flags);
- kfree(req->waitevent);
kfree(req);
}
@@ -321,7 +319,8 @@ static void rndis_filter_receive_response(struct rndis_device *dev,
}
}
- osd_waitevent_set(request->waitevent);
+ request->wait_condition = 1;
+ wake_up(&request->wait_event);
} else {
DPRINT_ERR(NETVSC, "no rndis request found for this response "
"(id 0x%x res type 0x%x)",
@@ -503,11 +502,17 @@ static int rndis_filter_query_device(struct rndis_device *dev, u32 oid,
query->info_buflen = 0;
query->dev_vc_handle = 0;
+ request->wait_condition = 0;
ret = rndis_filter_send_request(dev, request);
if (ret != 0)
goto Cleanup;
- osd_waitevent_wait(request->waitevent);
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto Cleanup;
+ }
/* Copy the response back */
query_complete = &request->response_msg.msg.query_complete;
@@ -578,12 +583,14 @@ static int rndis_filter_set_packet_filter(struct rndis_device *dev,
memcpy((void *)(unsigned long)set + sizeof(struct rndis_set_request),
&new_filter, sizeof(u32));
+ request->wait_condition = 0;
ret = rndis_filter_send_request(dev, request);
if (ret != 0)
goto Cleanup;
- ret = osd_waitevent_waitex(request->waitevent, 2000/*2sec*/);
- if (!ret) {
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(2000));
+ if (request->wait_condition == 0) {
ret = -1;
DPRINT_ERR(NETVSC, "timeout before we got a set response...");
/*
@@ -669,13 +676,20 @@ static int rndis_filter_init_device(struct rndis_device *dev)
dev->state = RNDIS_DEV_INITIALIZING;
+ request->wait_condition = 0;
ret = rndis_filter_send_request(dev, request);
if (ret != 0) {
dev->state = RNDIS_DEV_UNINITIALIZED;
goto Cleanup;
}
- osd_waitevent_wait(request->waitevent);
+
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto Cleanup;
+ }
init_complete = &request->response_msg.msg.init_complete;
status = init_complete->status;
diff --git a/drivers/staging/hv/storvsc.c b/drivers/staging/hv/storvsc.c
index a612109..2560342 100644
--- a/drivers/staging/hv/storvsc.c
+++ b/drivers/staging/hv/storvsc.c
@@ -19,6 +19,8 @@
* Hank Janssen <hjanssen@microsoft.com>
*/
#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/mm.h>
@@ -38,7 +40,8 @@ struct storvsc_request_extension {
struct hv_device *device;
/* Synchronize the request/response if needed */
- struct osd_waitevent *wait_event;
+ int wait_condition;
+ wait_queue_head_t wait_event;
struct vstor_packet vstor_packet;
};
@@ -200,21 +203,13 @@ static int stor_vsc_channel_init(struct hv_device *device)
* channel
*/
memset(request, 0, sizeof(struct storvsc_request_extension));
- request->wait_event = osd_waitevent_create();
- if (!request->wait_event) {
- ret = -ENOMEM;
- goto nomem;
- }
-
+ init_waitqueue_head(&request->wait_event);
vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
vstor_packet->flags = REQUEST_COMPLETION_FLAG;
- /*SpinlockAcquire(gDriverExt.packetListLock);
- INSERT_TAIL_LIST(&gDriverExt.packetList, &packet->listEntry.entry);
- SpinlockRelease(gDriverExt.packetListLock);*/
-
DPRINT_INFO(STORVSC, "BEGIN_INITIALIZATION_OPERATION...");
+ request->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, vstor_packet,
sizeof(struct vstor_packet),
(unsigned long)request,
@@ -223,17 +218,23 @@ static int stor_vsc_channel_init(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(STORVSC,
"unable to send BEGIN_INITIALIZATION_OPERATION");
- goto Cleanup;
+ goto cleanup;
+ }
+
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
}
- osd_waitevent_wait(request->wait_event);
if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
vstor_packet->status != 0) {
DPRINT_ERR(STORVSC, "BEGIN_INITIALIZATION_OPERATION failed "
"(op %d status 0x%x)",
vstor_packet->operation, vstor_packet->status);
- goto Cleanup;
+ goto cleanup;
}
DPRINT_INFO(STORVSC, "QUERY_PROTOCOL_VERSION_OPERATION...");
@@ -246,6 +247,7 @@ static int stor_vsc_channel_init(struct hv_device *device)
vstor_packet->version.major_minor = VMSTOR_PROTOCOL_VERSION_CURRENT;
FILL_VMSTOR_REVISION(vstor_packet->version.revision);
+ request->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, vstor_packet,
sizeof(struct vstor_packet),
(unsigned long)request,
@@ -254,10 +256,15 @@ static int stor_vsc_channel_init(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(STORVSC,
"unable to send BEGIN_INITIALIZATION_OPERATION");
- goto Cleanup;
+ goto cleanup;
}
- osd_waitevent_wait(request->wait_event);
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
/* TODO: Check returned version */
if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
@@ -265,7 +272,7 @@ static int stor_vsc_channel_init(struct hv_device *device)
DPRINT_ERR(STORVSC, "QUERY_PROTOCOL_VERSION_OPERATION failed "
"(op %d status 0x%x)",
vstor_packet->operation, vstor_packet->status);
- goto Cleanup;
+ goto cleanup;
}
/* Query channel properties */
@@ -277,6 +284,7 @@ static int stor_vsc_channel_init(struct hv_device *device)
vstor_packet->storage_channel_properties.port_number =
stor_device->port_number;
+ request->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, vstor_packet,
sizeof(struct vstor_packet),
(unsigned long)request,
@@ -286,10 +294,15 @@ static int stor_vsc_channel_init(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(STORVSC,
"unable to send QUERY_PROPERTIES_OPERATION");
- goto Cleanup;
+ goto cleanup;
}
- osd_waitevent_wait(request->wait_event);
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
/* TODO: Check returned version */
if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
@@ -297,7 +310,7 @@ static int stor_vsc_channel_init(struct hv_device *device)
DPRINT_ERR(STORVSC, "QUERY_PROPERTIES_OPERATION failed "
"(op %d status 0x%x)",
vstor_packet->operation, vstor_packet->status);
- goto Cleanup;
+ goto cleanup;
}
stor_device->path_id = vstor_packet->storage_channel_properties.path_id;
@@ -314,6 +327,7 @@ static int stor_vsc_channel_init(struct hv_device *device)
vstor_packet->operation = VSTOR_OPERATION_END_INITIALIZATION;
vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+ request->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, vstor_packet,
sizeof(struct vstor_packet),
(unsigned long)request,
@@ -323,25 +337,27 @@ static int stor_vsc_channel_init(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(STORVSC,
"unable to send END_INITIALIZATION_OPERATION");
- goto Cleanup;
+ goto cleanup;
}
- osd_waitevent_wait(request->wait_event);
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
vstor_packet->status != 0) {
DPRINT_ERR(STORVSC, "END_INITIALIZATION_OPERATION failed "
"(op %d status 0x%x)",
vstor_packet->operation, vstor_packet->status);
- goto Cleanup;
+ goto cleanup;
}
DPRINT_INFO(STORVSC, "**** storage channel up and running!! ****");
-Cleanup:
- kfree(request->wait_event);
- request->wait_event = NULL;
-nomem:
+cleanup:
put_stor_device(device);
return ret;
}
@@ -476,8 +492,8 @@ static void stor_vsc_on_channel_callback(void *context)
memcpy(&request->vstor_packet, packet,
sizeof(struct vstor_packet));
-
- osd_waitevent_set(request->wait_event);
+ request->wait_condition = 1;
+ wake_up(&request->wait_event);
} else {
stor_vsc_on_receive(device,
(struct vstor_packet *)packet,
@@ -539,7 +555,7 @@ static int stor_vsc_on_device_add(struct hv_device *device,
stor_device = alloc_stor_device(device);
if (!stor_device) {
ret = -1;
- goto Cleanup;
+ goto cleanup;
}
/* Save the channel properties to our storvsc channel */
@@ -569,7 +585,7 @@ static int stor_vsc_on_device_add(struct hv_device *device,
stor_device->port_number, stor_device->path_id,
stor_device->target_id);
-Cleanup:
+cleanup:
return ret;
}
@@ -629,16 +645,13 @@ int stor_vsc_on_host_reset(struct hv_device *device)
request = &stor_device->reset_request;
vstor_packet = &request->vstor_packet;
- request->wait_event = osd_waitevent_create();
- if (!request->wait_event) {
- ret = -ENOMEM;
- goto Cleanup;
- }
+ init_waitqueue_head(&request->wait_event);
vstor_packet->operation = VSTOR_OPERATION_RESET_BUS;
vstor_packet->flags = REQUEST_COMPLETION_FLAG;
vstor_packet->vm_srb.path_id = stor_device->path_id;
+ request->wait_condition = 0;
ret = vmbus_sendpacket(device->channel, vstor_packet,
sizeof(struct vstor_packet),
(unsigned long)&stor_device->reset_request,
@@ -647,13 +660,16 @@ int stor_vsc_on_host_reset(struct hv_device *device)
if (ret != 0) {
DPRINT_ERR(STORVSC, "Unable to send reset packet %p ret %d",
vstor_packet, ret);
- goto Cleanup;
+ goto cleanup;
}
- /* FIXME: Add a timeout */
- osd_waitevent_wait(request->wait_event);
+ wait_event_timeout(request->wait_event, request->wait_condition,
+ msecs_to_jiffies(1000));
+ if (request->wait_condition == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
- kfree(request->wait_event);
DPRINT_INFO(STORVSC, "host adapter reset completed");
/*
@@ -661,7 +677,7 @@ int stor_vsc_on_host_reset(struct hv_device *device)
* should have been flushed out and return to us
*/
-Cleanup:
+cleanup:
put_stor_device(device);
return ret;
}
diff --git a/drivers/staging/hv/vmbus_private.h b/drivers/staging/hv/vmbus_private.h
index 004d8de..9f505c4 100644
--- a/drivers/staging/hv/vmbus_private.h
+++ b/drivers/staging/hv/vmbus_private.h
@@ -91,7 +91,8 @@ struct vmbus_msginfo {
struct list_head msglist_entry;
/* Synchronize the request/response if needed */
- struct osd_waitevent *wait_event;
+ int wait_condition;
+ wait_queue_head_t wait_event;
/* The message itself */
unsigned char msg[0];
--
1.5.5.6
^ permalink raw reply related
* [PATCH 3/3]:Staging: hv: Remove osd layer
From: K. Y. Srinivasan @ 2011-02-11 18:00 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, virtualization; +Cc: K. Y. Srinivasan
The OSD layer was a wrapper around native interfaces
adding little value and was infact buggy -
refer to the osd_wait.patch for details.
This patch gets rid of the OSD abstraction.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Hank Janssen <hjanssen@microsoft.com>
---
drivers/staging/hv/Makefile | 2 +-
drivers/staging/hv/blkvsc.c | 2 +-
drivers/staging/hv/blkvsc_drv.c | 2 +-
drivers/staging/hv/channel.c | 5 +-
drivers/staging/hv/channel_mgmt.c | 2 +-
drivers/staging/hv/connection.c | 2 +-
drivers/staging/hv/hv.c | 2 +-
drivers/staging/hv/hv_api.h | 5 +
drivers/staging/hv/hv_kvp.c | 2 +-
drivers/staging/hv/hv_util.c | 2 +-
drivers/staging/hv/logging.h | 3 +
drivers/staging/hv/netvsc.c | 2 +-
drivers/staging/hv/netvsc_drv.c | 2 +-
drivers/staging/hv/osd.c | 194 -------------------------------------
drivers/staging/hv/osd.h | 62 ------------
drivers/staging/hv/ring_buffer.c | 1 -
drivers/staging/hv/rndis_filter.c | 2 +-
drivers/staging/hv/storvsc.c | 2 +-
drivers/staging/hv/storvsc_drv.c | 2 +-
drivers/staging/hv/vmbus_drv.c | 2 +-
20 files changed, 26 insertions(+), 272 deletions(-)
delete mode 100644 drivers/staging/hv/osd.c
delete mode 100644 drivers/staging/hv/osd.h
diff --git a/drivers/staging/hv/Makefile b/drivers/staging/hv/Makefile
index 606ce7d..737e517 100644
--- a/drivers/staging/hv/Makefile
+++ b/drivers/staging/hv/Makefile
@@ -4,7 +4,7 @@ obj-$(CONFIG_HYPERV_BLOCK) += hv_blkvsc.o
obj-$(CONFIG_HYPERV_NET) += hv_netvsc.o
obj-$(CONFIG_HYPERV_UTILS) += hv_utils.o
-hv_vmbus-y := vmbus_drv.o osd.o \
+hv_vmbus-y := vmbus_drv.o \
hv.o connection.o channel.o \
channel_mgmt.o ring_buffer.o
hv_storvsc-y := storvsc_drv.o storvsc.o
diff --git a/drivers/staging/hv/blkvsc.c b/drivers/staging/hv/blkvsc.c
index b0e07c1..7c8729b 100644
--- a/drivers/staging/hv/blkvsc.c
+++ b/drivers/staging/hv/blkvsc.c
@@ -22,7 +22,7 @@
*/
#include <linux/kernel.h>
#include <linux/mm.h>
-#include "osd.h"
+#include "hv_api.h"
#include "storvsc.c"
static const char *g_blk_driver_name = "blkvsc";
diff --git a/drivers/staging/hv/blkvsc_drv.c b/drivers/staging/hv/blkvsc_drv.c
index 58bbcd6..36a0adb 100644
--- a/drivers/staging/hv/blkvsc_drv.c
+++ b/drivers/staging/hv/blkvsc_drv.c
@@ -31,7 +31,7 @@
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_eh.h>
#include <scsi/scsi_dbg.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "version_info.h"
#include "vmbus.h"
diff --git a/drivers/staging/hv/channel.c b/drivers/staging/hv/channel.c
index 5a0923c..775a52a 100644
--- a/drivers/staging/hv/channel.c
+++ b/drivers/staging/hv/channel.c
@@ -24,10 +24,13 @@
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/module.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "vmbus_private.h"
+#define NUM_PAGES_SPANNED(addr, len) \
+((PAGE_ALIGN(addr + len) >> PAGE_SHIFT) - (addr >> PAGE_SHIFT))
+
/* Internal routines */
static int create_gpadl_header(
void *kbuffer, /* must be phys and virt contiguous */
diff --git a/drivers/staging/hv/channel_mgmt.c b/drivers/staging/hv/channel_mgmt.c
index da1e56a..325e0bc 100644
--- a/drivers/staging/hv/channel_mgmt.c
+++ b/drivers/staging/hv/channel_mgmt.c
@@ -26,7 +26,7 @@
#include <linux/list.h>
#include <linux/module.h>
#include <linux/completion.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "vmbus_private.h"
#include "utils.h"
diff --git a/drivers/staging/hv/connection.c b/drivers/staging/hv/connection.c
index 51dd362..f7df479 100644
--- a/drivers/staging/hv/connection.c
+++ b/drivers/staging/hv/connection.c
@@ -26,7 +26,7 @@
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "vmbus_private.h"
diff --git a/drivers/staging/hv/hv.c b/drivers/staging/hv/hv.c
index 419b4d6..31b9073 100644
--- a/drivers/staging/hv/hv.c
+++ b/drivers/staging/hv/hv.c
@@ -23,7 +23,7 @@
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "vmbus_private.h"
diff --git a/drivers/staging/hv/hv_api.h b/drivers/staging/hv/hv_api.h
index 70e863a..7114fce 100644
--- a/drivers/staging/hv/hv_api.h
+++ b/drivers/staging/hv/hv_api.h
@@ -23,6 +23,11 @@
#ifndef __HV_API_H
#define __HV_API_H
+struct hv_guid {
+ unsigned char data[16];
+};
+
+
/* Status codes for hypervisor operations. */
diff --git a/drivers/staging/hv/hv_kvp.c b/drivers/staging/hv/hv_kvp.c
index bc1c20e..faf692e 100644
--- a/drivers/staging/hv/hv_kvp.c
+++ b/drivers/staging/hv/hv_kvp.c
@@ -28,7 +28,7 @@
#include <linux/workqueue.h>
#include "logging.h"
-#include "osd.h"
+#include "hv_api.h"
#include "vmbus.h"
#include "vmbus_packet_format.h"
#include "vmbus_channel_interface.h"
diff --git a/drivers/staging/hv/hv_util.c b/drivers/staging/hv/hv_util.c
index 43c7ec0..4792f2c 100644
--- a/drivers/staging/hv/hv_util.c
+++ b/drivers/staging/hv/hv_util.c
@@ -28,7 +28,7 @@
#include <linux/pci.h>
#include "logging.h"
-#include "osd.h"
+#include "hv_api.h"
#include "vmbus.h"
#include "vmbus_packet_format.h"
#include "vmbus_channel_interface.h"
diff --git a/drivers/staging/hv/logging.h b/drivers/staging/hv/logging.h
index 20d4d12..1799951 100644
--- a/drivers/staging/hv/logging.h
+++ b/drivers/staging/hv/logging.h
@@ -25,6 +25,9 @@
#ifndef _LOGGING_H_
#define _LOGGING_H_
+#define LOWORD(dw) ((unsigned short)(dw))
+#define HIWORD(dw) ((unsigned short)(((unsigned int) (dw) >> 16) & 0xFFFF))
+
/* #include <linux/init.h> */
/* #include <linux/module.h> */
diff --git a/drivers/staging/hv/netvsc.c b/drivers/staging/hv/netvsc.c
index 7233564..fa46a7e 100644
--- a/drivers/staging/hv/netvsc.c
+++ b/drivers/staging/hv/netvsc.c
@@ -25,7 +25,7 @@
#include <linux/delay.h>
#include <linux/io.h>
#include <linux/slab.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "netvsc.h"
#include "rndis_filter.h"
diff --git a/drivers/staging/hv/netvsc_drv.c b/drivers/staging/hv/netvsc_drv.c
index 6b8fd0c..e190738 100644
--- a/drivers/staging/hv/netvsc_drv.c
+++ b/drivers/staging/hv/netvsc_drv.c
@@ -36,7 +36,7 @@
#include <net/route.h>
#include <net/sock.h>
#include <net/pkt_sched.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "version_info.h"
#include "vmbus.h"
diff --git a/drivers/staging/hv/osd.c b/drivers/staging/hv/osd.c
deleted file mode 100644
index b5a3940..0000000
--- a/drivers/staging/hv/osd.c
+++ /dev/null
@@ -1,194 +0,0 @@
-/*
- *
- * Copyright (c) 2009, Microsoft Corporation.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59 Temple
- * Place - Suite 330, Boston, MA 02111-1307 USA.
- *
- * Authors:
- * Haiyang Zhang <haiyangz@microsoft.com>
- * Hank Janssen <hjanssen@microsoft.com>
- *
- */
-
-#include <linux/module.h>
-#include <linux/init.h>
-#include <linux/types.h>
-#include <linux/mm.h>
-#include <linux/highmem.h>
-#include <linux/vmalloc.h>
-#include <linux/ioport.h>
-#include <linux/irq.h>
-#include <linux/interrupt.h>
-#include <linux/sched.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/workqueue.h>
-#include <linux/kernel.h>
-#include <linux/jiffies.h>
-#include <linux/delay.h>
-#include <linux/time.h>
-#include <linux/io.h>
-#include <linux/bitops.h>
-#include <linux/slab.h>
-#include "osd.h"
-
-void *osd_virtual_alloc_exec(unsigned int size)
-{
-#ifdef __x86_64__
- return __vmalloc(size, GFP_KERNEL, PAGE_KERNEL_EXEC);
-#else
- return __vmalloc(size, GFP_KERNEL,
- __pgprot(__PAGE_KERNEL & (~_PAGE_NX)));
-#endif
-}
-
-/**
- * osd_page_alloc() - Allocate pages
- * @count: Total number of Kernel pages you want to allocate
- *
- * Tries to allocate @count number of consecutive free kernel pages.
- * And if successful, it will set the pages to 0 before returning.
- * If successfull it will return pointer to the @count pages.
- * Mainly used by Hyper-V drivers.
- */
-void *osd_page_alloc(unsigned int count)
-{
- void *p;
-
- p = (void *)__get_free_pages(GFP_KERNEL, get_order(count * PAGE_SIZE));
- if (p)
- memset(p, 0, count * PAGE_SIZE);
- return p;
-
- /* struct page* page = alloc_page(GFP_KERNEL|__GFP_ZERO); */
- /* void *p; */
-
- /* BUGBUG: We need to use kmap in case we are in HIMEM region */
- /* p = page_address(page); */
- /* if (p) memset(p, 0, PAGE_SIZE); */
- /* return p; */
-}
-EXPORT_SYMBOL_GPL(osd_page_alloc);
-
-/**
- * osd_page_free() - Free pages
- * @page: Pointer to the first page to be freed
- * @count: Total number of Kernel pages you free
- *
- * Frees the pages allocated by osd_page_alloc()
- * Mainly used by Hyper-V drivers.
- */
-void osd_page_free(void *page, unsigned int count)
-{
- free_pages((unsigned long)page, get_order(count * PAGE_SIZE));
- /*struct page* p = virt_to_page(page);
- __free_page(p);*/
-}
-EXPORT_SYMBOL_GPL(osd_page_free);
-
-/**
- * osd_waitevent_create() - Create the event queue
- *
- * Allocates memory for a &struct osd_waitevent. And then calls
- * init_waitqueue_head to set up the wait queue for the event.
- * This structure is usually part of a another structure that contains
- * the actual Hyper-V device driver structure.
- *
- * Returns pointer to &struct osd_waitevent
- * Mainly used by Hyper-V drivers.
- */
-struct osd_waitevent *osd_waitevent_create(void)
-{
- struct osd_waitevent *wait = kmalloc(sizeof(struct osd_waitevent),
- GFP_KERNEL);
- if (!wait)
- return NULL;
-
- wait->condition = 0;
- init_waitqueue_head(&wait->event);
- return wait;
-}
-EXPORT_SYMBOL_GPL(osd_waitevent_create);
-
-
-/**
- * osd_waitevent_set() - Wake up the process
- * @wait_event: Structure to event to be woken up
- *
- * @wait_event is of type &struct osd_waitevent
- *
- * Wake up the sleeping process so it can do some work.
- * And set condition indicator in &struct osd_waitevent to indicate
- * the process is in a woken state.
- *
- * Only used by Network and Storage Hyper-V drivers.
- */
-void osd_waitevent_set(struct osd_waitevent *wait_event)
-{
- wait_event->condition = 1;
- wake_up_interruptible(&wait_event->event);
-}
-EXPORT_SYMBOL_GPL(osd_waitevent_set);
-
-/**
- * osd_waitevent_wait() - Wait for event till condition is true
- * @wait_event: Structure to event to be put to sleep
- *
- * @wait_event is of type &struct osd_waitevent
- *
- * Set up the process to sleep until waitEvent->condition get true.
- * And set condition indicator in &struct osd_waitevent to indicate
- * the process is in a sleeping state.
- *
- * Returns the status of 'wait_event_interruptible()' system call
- *
- * Mainly used by Hyper-V drivers.
- */
-int osd_waitevent_wait(struct osd_waitevent *wait_event)
-{
- int ret = 0;
-
- ret = wait_event_interruptible(wait_event->event,
- wait_event->condition);
- wait_event->condition = 0;
- return ret;
-}
-EXPORT_SYMBOL_GPL(osd_waitevent_wait);
-
-/**
- * osd_waitevent_waitex() - Wait for event or timeout for process wakeup
- * @wait_event: Structure to event to be put to sleep
- * @timeout_in_ms: Total number of Milliseconds to wait before waking up
- *
- * @wait_event is of type &struct osd_waitevent
- * Set up the process to sleep until @waitEvent->condition get true or
- * @timeout_in_ms (Time out in Milliseconds) has been reached.
- * And set condition indicator in &struct osd_waitevent to indicate
- * the process is in a sleeping state.
- *
- * Returns the status of 'wait_event_interruptible_timeout()' system call
- *
- * Mainly used by Hyper-V drivers.
- */
-int osd_waitevent_waitex(struct osd_waitevent *wait_event, u32 timeout_in_ms)
-{
- int ret = 0;
-
- ret = wait_event_interruptible_timeout(wait_event->event,
- wait_event->condition,
- msecs_to_jiffies(timeout_in_ms));
- wait_event->condition = 0;
- return ret;
-}
-EXPORT_SYMBOL_GPL(osd_waitevent_waitex);
diff --git a/drivers/staging/hv/osd.h b/drivers/staging/hv/osd.h
deleted file mode 100644
index f787161..0000000
--- a/drivers/staging/hv/osd.h
+++ /dev/null
@@ -1,62 +0,0 @@
-/*
- *
- * Copyright (c) 2009, Microsoft Corporation.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59 Temple
- * Place - Suite 330, Boston, MA 02111-1307 USA.
- *
- * Authors:
- * Haiyang Zhang <haiyangz@microsoft.com>
- * Hank Janssen <hjanssen@microsoft.com>
- *
- */
-
-
-#ifndef _OSD_H_
-#define _OSD_H_
-
-#include <linux/kernel.h>
-#include <linux/workqueue.h>
-
-/* Defines */
-#define NUM_PAGES_SPANNED(addr, len) ((PAGE_ALIGN(addr + len) >> PAGE_SHIFT) - \
- (addr >> PAGE_SHIFT))
-
-#define LOWORD(dw) ((unsigned short)(dw))
-#define HIWORD(dw) ((unsigned short)(((unsigned int) (dw) >> 16) & 0xFFFF))
-
-struct hv_guid {
- unsigned char data[16];
-};
-
-struct osd_waitevent {
- int condition;
- wait_queue_head_t event;
-};
-
-/* Osd routines */
-
-extern void *osd_virtual_alloc_exec(unsigned int size);
-
-extern void *osd_page_alloc(unsigned int count);
-extern void osd_page_free(void *page, unsigned int count);
-
-extern struct osd_waitevent *osd_waitevent_create(void);
-extern void osd_waitevent_set(struct osd_waitevent *wait_event);
-extern int osd_waitevent_wait(struct osd_waitevent *wait_event);
-
-/* If >0, wait_event got signaled. If ==0, timeout. If < 0, error */
-extern int osd_waitevent_waitex(struct osd_waitevent *wait_event,
- u32 timeout_in_ms);
-
-#endif /* _OSD_H_ */
diff --git a/drivers/staging/hv/ring_buffer.c b/drivers/staging/hv/ring_buffer.c
index 4d53392..66688fb 100644
--- a/drivers/staging/hv/ring_buffer.c
+++ b/drivers/staging/hv/ring_buffer.c
@@ -23,7 +23,6 @@
#include <linux/kernel.h>
#include <linux/mm.h>
-#include "osd.h"
#include "logging.h"
#include "ring_buffer.h"
diff --git a/drivers/staging/hv/rndis_filter.c b/drivers/staging/hv/rndis_filter.c
index e3bf004..9dde936 100644
--- a/drivers/staging/hv/rndis_filter.c
+++ b/drivers/staging/hv/rndis_filter.c
@@ -26,8 +26,8 @@
#include <linux/io.h>
#include <linux/if_ether.h>
-#include "osd.h"
#include "logging.h"
+#include "hv_api.h"
#include "netvsc_api.h"
#include "rndis_filter.h"
diff --git a/drivers/staging/hv/storvsc.c b/drivers/staging/hv/storvsc.c
index 2560342..e2ad729 100644
--- a/drivers/staging/hv/storvsc.c
+++ b/drivers/staging/hv/storvsc.c
@@ -25,7 +25,7 @@
#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/delay.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "storvsc_api.h"
#include "vmbus_packet_format.h"
diff --git a/drivers/staging/hv/storvsc_drv.c b/drivers/staging/hv/storvsc_drv.c
index 956c9eb..a8427ff 100644
--- a/drivers/staging/hv/storvsc_drv.c
+++ b/drivers/staging/hv/storvsc_drv.c
@@ -31,7 +31,7 @@
#include <scsi/scsi_eh.h>
#include <scsi/scsi_devinfo.h>
#include <scsi/scsi_dbg.h>
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "version_info.h"
#include "vmbus.h"
diff --git a/drivers/staging/hv/vmbus_drv.c b/drivers/staging/hv/vmbus_drv.c
index dacaa54..459c707 100644
--- a/drivers/staging/hv/vmbus_drv.c
+++ b/drivers/staging/hv/vmbus_drv.c
@@ -29,7 +29,7 @@
#include <linux/slab.h>
#include <linux/completion.h>
#include "version_info.h"
-#include "osd.h"
+#include "hv_api.h"
#include "logging.h"
#include "vmbus.h"
#include "channel.h"
--
1.5.5.6
^ permalink raw reply related
* Re: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: Greg KH @ 2011-02-11 18:29 UTC (permalink / raw)
To: K. Y. Srinivasan; +Cc: linux-kernel, devel, virtualization, Hank Janssen
In-Reply-To: <1297447140-21777-1-git-send-email-kys@microsoft.com>
On Fri, Feb 11, 2011 at 09:59:00AM -0800, K. Y. Srinivasan wrote:
> --- a/drivers/staging/hv/hv.c
> +++ b/drivers/staging/hv/hv.c
> @@ -230,7 +230,12 @@ int hv_init(void)
> * Allocate the hypercall page memory
> * virtaddr = osd_page_alloc(1);
> */
> - virtaddr = osd_virtual_alloc_exec(PAGE_SIZE);
> +#ifdef __x86_64__
> + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC);
> +#else
> + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
> + __pgprot(__PAGE_KERNEL & (~_PAGE_NX)));
> +#endif
I'm not saying this patch is wrong at all, but I still don't understand
why this is different depending on the architecture of the machine. Why
is this necessary, it should be ok to do the same type of allocation no
matter what the processor is, right?
thanks,
greg k-h
^ permalink raw reply
* RE: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: KY Srinivasan @ 2011-02-11 20:55 UTC (permalink / raw)
To: Greg KH
Cc: devel@linuxdriverproject.org, linux-kernel@vger.kernel.org,
virtualization@lists.osdl.org
In-Reply-To: <20110211182949.GA10751@suse.de>
> -----Original Message-----
> From: Greg KH [mailto:gregkh@suse.de]
> Sent: Friday, February 11, 2011 1:30 PM
> To: KY Srinivasan
> Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org;
> virtualization@lists.osdl.org; Hank Janssen
> Subject: Re: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
>
> On Fri, Feb 11, 2011 at 09:59:00AM -0800, K. Y. Srinivasan wrote:
> > --- a/drivers/staging/hv/hv.c
> > +++ b/drivers/staging/hv/hv.c
> > @@ -230,7 +230,12 @@ int hv_init(void)
> > * Allocate the hypercall page memory
> > * virtaddr = osd_page_alloc(1);
> > */
> > - virtaddr = osd_virtual_alloc_exec(PAGE_SIZE);
> > +#ifdef __x86_64__
> > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC);
> #else
> > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
> > + __pgprot(__PAGE_KERNEL & (~_PAGE_NX))); #endif
>
> I'm not saying this patch is wrong at all, but I still don't understand why this is
> different depending on the architecture of the machine. Why is this necessary, it
> should be ok to do the same type of allocation no matter what the processor is,
> right?
You are right Greg; I don't think there is a need to specify different page
protection bits based on the architecture - PAGE_KERNEL_EXEC should be enough.
However, this is the code that is currently in the tree - refer to osd.c.
If it is ok with you, I could submit an additional patch to clean this up.
Regards,
K. Y
>
> thanks,
>
> greg k-h
^ permalink raw reply
* Re: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: Greg KH @ 2011-02-11 21:23 UTC (permalink / raw)
To: KY Srinivasan
Cc: devel@linuxdriverproject.org, linux-kernel@vger.kernel.org,
virtualization@lists.osdl.org
In-Reply-To: <FB42D5CCD7B5934EB1827DB5ED9B850E0B2653@tk5ex14mbxc106.redmond.corp.microsoft.com>
On Fri, Feb 11, 2011 at 08:55:56PM +0000, KY Srinivasan wrote:
>
>
> > -----Original Message-----
> > From: Greg KH [mailto:gregkh@suse.de]
> > Sent: Friday, February 11, 2011 1:30 PM
> > To: KY Srinivasan
> > Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org;
> > virtualization@lists.osdl.org; Hank Janssen
> > Subject: Re: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
> >
> > On Fri, Feb 11, 2011 at 09:59:00AM -0800, K. Y. Srinivasan wrote:
> > > --- a/drivers/staging/hv/hv.c
> > > +++ b/drivers/staging/hv/hv.c
> > > @@ -230,7 +230,12 @@ int hv_init(void)
> > > * Allocate the hypercall page memory
> > > * virtaddr = osd_page_alloc(1);
> > > */
> > > - virtaddr = osd_virtual_alloc_exec(PAGE_SIZE);
> > > +#ifdef __x86_64__
> > > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC);
> > #else
> > > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
> > > + __pgprot(__PAGE_KERNEL & (~_PAGE_NX))); #endif
> >
> > I'm not saying this patch is wrong at all, but I still don't understand why this is
> > different depending on the architecture of the machine. Why is this necessary, it
> > should be ok to do the same type of allocation no matter what the processor is,
> > right?
>
> You are right Greg; I don't think there is a need to specify different page
> protection bits based on the architecture - PAGE_KERNEL_EXEC should be enough.
I thought so, but for some reason Hank said there this was needed.
Hank, is it still true?
> However, this is the code that is currently in the tree - refer to osd.c.
Oh, I remember, it's not a critique of this patch, it just reminded me
of this question I always had for this code.
> If it is ok with you, I could submit an additional patch to clean this up.
If Hank says it is ok, and you all test it to verify nothing breaks,
please send it on.
thanks,
greg k-h
^ permalink raw reply
* RE: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: Hank Janssen @ 2011-02-11 21:27 UTC (permalink / raw)
To: KY Srinivasan, Greg KH
Cc: linux-kernel@vger.kernel.org, devel@linuxdriverproject.org,
virtualization@lists.osdl.org
In-Reply-To: <FB42D5CCD7B5934EB1827DB5ED9B850E0B2653@tk5ex14mbxc106.redmond.corp.microsoft.com>
> -----Original Message-----
> And then KY Srinivasan spoke on Friday, February 11, 2011 12:56 PM
> > From: Greg KH [mailto:gregkh@suse.de]
> > Sent: Friday, February 11, 2011 1:30 PM
> > > - virtaddr = osd_virtual_alloc_exec(PAGE_SIZE);
> > > +#ifdef __x86_64__
> > > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
> PAGE_KERNEL_EXEC);
> > #else
> > > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
> > > + __pgprot(__PAGE_KERNEL & (~_PAGE_NX))); #endif
> >
> > I'm not saying this patch is wrong at all, but I still don't
> > understand why this is different depending on the architecture of the
> > machine. Why is this necessary, it should be ok to do the same type
> > of allocation no matter what the processor is, right?
>
> You are right Greg; I don't think there is a need to specify different page
> protection bits based on the architecture - PAGE_KERNEL_EXEC should be
> enough.
> However, this is the code that is currently in the tree - refer to osd.c.
> If it is ok with you, I could submit an additional patch to clean this up.
>
I seem to recall that we did it for very old versions of Linux (pre 2.6.18)
But I cannot for the life of me remember why.
Hank.
^ permalink raw reply
* RE: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: Hank Janssen @ 2011-02-11 21:30 UTC (permalink / raw)
To: Greg KH, KY Srinivasan
Cc: linux-kernel@vger.kernel.org, devel@linuxdriverproject.org,
virtualization@lists.osdl.org
In-Reply-To: <20110211212352.GA10845@suse.de>
> -----Original Message-----
> From: Greg KH [mailto:gregkh@suse.de] On Friday, February 11, 2011 1:24 PM
> On Fri, Feb 11, 2011 at 08:55:56PM +0000, KY Srinivasan wrote:
> > > I'm not saying this patch is wrong at all, but I still don't
> > > understand why this is different depending on the architecture of
> > > the machine. Why is this necessary, it should be ok to do the same
> > > type of allocation no matter what the processor is, right?
> >
> > You are right Greg; I don't think there is a need to specify different
> > page protection bits based on the architecture - PAGE_KERNEL_EXEC
> should be enough.
>
> I thought so, but for some reason Hank said there this was needed.
> Hank, is it still true?
I recall we did it for older versions of Linux but I do not recall why.
Something from way before 2.6.18, the reason of which I seem to have
Purged due to age :)
>
> > However, this is the code that is currently in the tree - refer to osd.c.
>
> Oh, I remember, it's not a critique of this patch, it just reminded me of this
> question I always had for this code.
>
> > If it is ok with you, I could submit an additional patch to clean this up.
>
> If Hank says it is ok, and you all test it to verify nothing breaks, please send it
> on.
If you could accept the patch as is and I will work with Ky to see if nothing breaks
If we change this part to what you are suggesting? If nothing breaks we will submit
A followup patch to remove those lines.
Hank.
^ permalink raw reply
* RE: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
From: KY Srinivasan @ 2011-02-11 21:37 UTC (permalink / raw)
To: Greg KH
Cc: devel@linuxdriverproject.org, linux-kernel@vger.kernel.org,
virtualization@lists.osdl.org
In-Reply-To: <20110211212352.GA10845@suse.de>
> -----Original Message-----
> From: Greg KH [mailto:gregkh@suse.de]
> Sent: Friday, February 11, 2011 4:24 PM
> To: KY Srinivasan
> Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org;
> virtualization@lists.osdl.org; Hank Janssen
> Subject: Re: [PATCH 1/3]: Staging: hv: Use native page allocation/free functions
>
> On Fri, Feb 11, 2011 at 08:55:56PM +0000, KY Srinivasan wrote:
> >
> >
> > > -----Original Message-----
> > > From: Greg KH [mailto:gregkh@suse.de]
> > > Sent: Friday, February 11, 2011 1:30 PM
> > > To: KY Srinivasan
> > > Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org;
> > > virtualization@lists.osdl.org; Hank Janssen
> > > Subject: Re: [PATCH 1/3]: Staging: hv: Use native page allocation/free
> functions
> > >
> > > On Fri, Feb 11, 2011 at 09:59:00AM -0800, K. Y. Srinivasan wrote:
> > > > --- a/drivers/staging/hv/hv.c
> > > > +++ b/drivers/staging/hv/hv.c
> > > > @@ -230,7 +230,12 @@ int hv_init(void)
> > > > * Allocate the hypercall page memory
> > > > * virtaddr = osd_page_alloc(1);
> > > > */
> > > > - virtaddr = osd_virtual_alloc_exec(PAGE_SIZE);
> > > > +#ifdef __x86_64__
> > > > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC);
> > > #else
> > > > + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
> > > > + __pgprot(__PAGE_KERNEL & (~_PAGE_NX))); #endif
> > >
> > > I'm not saying this patch is wrong at all, but I still don't understand why this is
> > > different depending on the architecture of the machine. Why is this
> necessary, it
> > > should be ok to do the same type of allocation no matter what the processor
> is,
> > > right?
> >
> > You are right Greg; I don't think there is a need to specify different page
> > protection bits based on the architecture - PAGE_KERNEL_EXEC should be
> enough.
>
> I thought so, but for some reason Hank said there this was needed.
> Hank, is it still true?
>
> > However, this is the code that is currently in the tree - refer to osd.c.
>
> Oh, I remember, it's not a critique of this patch, it just reminded me
> of this question I always had for this code.
>
> > If it is ok with you, I could submit an additional patch to clean this up.
>
> If Hank says it is ok, and you all test it to verify nothing breaks,
> please send it on.
Will do; the new patch will be on top of the 3 patch sequence I sent earlier today.
Regards,
K. Y
>
> thanks,
>
> greg k-h
^ permalink raw reply
* [PATCH]: Staging: hv: Cleanup vmalloc calls
From: K. Y. Srinivasan @ 2011-02-11 22:48 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, virtualization
Cc: K. Y. Srinivasan, Hank Janssen
The subject says it all. There is no need to
specify different page protection bits based
on the architecture.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Hank Janssen <hjanssen@microsoft.com>
---
drivers/staging/hv/hv.c | 5 -----
1 files changed, 0 insertions(+), 5 deletions(-)
diff --git a/drivers/staging/hv/hv.c b/drivers/staging/hv/hv.c
index 31b9073..2d492ad 100644
--- a/drivers/staging/hv/hv.c
+++ b/drivers/staging/hv/hv.c
@@ -230,12 +230,7 @@ int hv_init(void)
* Allocate the hypercall page memory
* virtaddr = osd_page_alloc(1);
*/
-#ifdef __x86_64__
virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC);
-#else
- virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL,
- __pgprot(__PAGE_KERNEL & (~_PAGE_NX)));
-#endif
if (!virtaddr) {
DPRINT_ERR(VMBUS,
--
1.5.5.6
^ permalink raw reply related
* Re: [PATCH]: Staging: hv: Cleanup vmalloc calls
From: Greg KH @ 2011-02-11 22:53 UTC (permalink / raw)
To: K. Y. Srinivasan; +Cc: linux-kernel, devel, virtualization, Hank Janssen
In-Reply-To: <1297464525-23254-1-git-send-email-kys@microsoft.com>
On Fri, Feb 11, 2011 at 02:48:45PM -0800, K. Y. Srinivasan wrote:
> The subject says it all. There is no need to
> specify different page protection bits based
> on the architecture.
Wonderful, thanks for doing this. I'll queue it up after your previous
patches sometime next week.
thanks,
greg k-h
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox