* Virtio_console usage of early printk
@ 2010-03-22 17:45 François Diakhate
2010-03-23 2:16 ` Amit Shah
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: François Diakhate @ 2010-03-22 17:45 UTC (permalink / raw)
To: virtualization, linuxppc-dev; +Cc: amit.shah, fdiakh
[-- Attachment #1: Type: text/plain, Size: 449 bytes --]
Hi all,
As far as I can see, early_put_chars is not used by virtio_console
because it checks whether there is a port available before using it.
If I understand correctly, this makes it useless because once we have
a port, we can use the regular virtio transport to output things to
the console. Does the attached patch seem valid ? Feedback from s390
and powerpc users who use this functionality would be appreciated.
Thanks,
François
[-- Attachment #2: 0001-virtio-console-Fix-early_put_chars-usage.patch --]
[-- Type: text/x-patch, Size: 1112 bytes --]
From 3961f380bbe84a1036ddfc823039cbee31b44dcb Mon Sep 17 00:00:00 2001
From: =?utf-8?q?Fran=C3=A7ois=20Diakhat=C3=A9?= <fdiakh@gmail.com>
Date: Thu, 18 Mar 2010 14:48:20 +0100
Subject: virtio: console: Fix early_put_chars usage
Currently early_put_chars is not used by virtio_console because it can only be used once a port has been found, at which point it's too late because it is no longer needed. This patch should fix it.
---
drivers/char/virtio_console.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
index f404ccf..691ba21 100644
--- a/drivers/char/virtio_console.c
+++ b/drivers/char/virtio_console.c
@@ -645,13 +645,13 @@ static int put_chars(u32 vtermno, const char *buf, int count)
{
struct port *port;
+ if (unlikely(early_put_chars))
+ return early_put_chars(vtermno, buf, count);
+
port = find_port_by_vtermno(vtermno);
if (!port)
return 0;
- if (unlikely(early_put_chars))
- return early_put_chars(vtermno, buf, count);
-
return send_buf(port, (void *)buf, count);
}
--
1.6.1.3
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: Virtio_console usage of early printk
2010-03-22 17:45 Virtio_console usage of early printk François Diakhate
@ 2010-03-23 2:16 ` Amit Shah
2010-03-23 8:17 ` Christian Borntraeger
2010-03-23 7:42 ` Alexander Graf
2010-03-23 8:55 ` Amit Shah
2 siblings, 1 reply; 6+ messages in thread
From: Amit Shah @ 2010-03-23 2:16 UTC (permalink / raw)
To: François Diakhate
Cc: Christian Borntraeger, linuxppc-dev, virtualization
On (Mon) Mar 22 2010 [18:45:47], François Diakhate wrote:
> Hi all,
>
> As far as I can see, early_put_chars is not used by virtio_console
> because it checks whether there is a port available before using it.
> If I understand correctly, this makes it useless because once we have
> a port, we can use the regular virtio transport to output things to
> the console. Does the attached patch seem valid ? Feedback from s390
> and powerpc users who use this functionality would be appreciated.
Looks OK to me, Christian, others, could you comment on this please?
> Thanks,
> François
> From 3961f380bbe84a1036ddfc823039cbee31b44dcb Mon Sep 17 00:00:00 2001
> From: =?utf-8?q?Fran=C3=A7ois=20Diakhat=C3=A9?= <fdiakh@gmail.com>
> Date: Thu, 18 Mar 2010 14:48:20 +0100
> Subject: virtio: console: Fix early_put_chars usage
>
> Currently early_put_chars is not used by virtio_console because it can only be used once a port has been found, at which point it's too late because it is no longer needed. This patch should fix it.
> ---
> drivers/char/virtio_console.c | 6 +++---
> 1 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
> index f404ccf..691ba21 100644
> --- a/drivers/char/virtio_console.c
> +++ b/drivers/char/virtio_console.c
> @@ -645,13 +645,13 @@ static int put_chars(u32 vtermno, const char *buf, int count)
> {
> struct port *port;
>
> + if (unlikely(early_put_chars))
> + return early_put_chars(vtermno, buf, count);
> +
> port = find_port_by_vtermno(vtermno);
> if (!port)
> return 0;
>
> - if (unlikely(early_put_chars))
> - return early_put_chars(vtermno, buf, count);
> -
> return send_buf(port, (void *)buf, count);
> }
>
> --
> 1.6.1.3
>
Amit
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Virtio_console usage of early printk
2010-03-22 17:45 Virtio_console usage of early printk François Diakhate
2010-03-23 2:16 ` Amit Shah
@ 2010-03-23 7:42 ` Alexander Graf
2010-03-23 8:55 ` Amit Shah
2 siblings, 0 replies; 6+ messages in thread
From: Alexander Graf @ 2010-03-23 7:42 UTC (permalink / raw)
To: François Diakhate; +Cc: amit.shah, linuxppc-dev, virtualization
On 22.03.2010, at 18:45, Fran=E7ois Diakhate wrote:
> Hi all,
>=20
> As far as I can see, early_put_chars is not used by virtio_console
> because it checks whether there is a port available before using it.
> If I understand correctly, this makes it useless because once we have
> a port, we can use the regular virtio transport to output things to
> the console. Does the attached patch seem valid ? Feedback from s390
> and powerpc users who use this functionality would be appreciated.
On S390 we use the hypercall interface with characters instead of memory =
addresses as parameters. The hypervisor then decides if the character is =
an early printk or a real hypercall.
I'm not sure how PPC handles that - I never used virtio console there.
Either way, early printk does work for S390. So what exactly is this =
patch supposed to fix?
Alex=
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Virtio_console usage of early printk
2010-03-23 2:16 ` Amit Shah
@ 2010-03-23 8:17 ` Christian Borntraeger
0 siblings, 0 replies; 6+ messages in thread
From: Christian Borntraeger @ 2010-03-23 8:17 UTC (permalink / raw)
To: Amit Shah; +Cc: François Diakhate, linuxppc-dev, virtualization
Am Dienstag 23 M=E4rz 2010 03:16:20 schrieb Amit Shah:
> On (Mon) Mar 22 2010 [18:45:47], Fran=E7ois Diakhate wrote:
> > the console. Does the attached patch seem valid ? Feedback from s390
> > and powerpc users who use this functionality would be appreciated.
>=20
> Looks OK to me, Christian, others, could you comment on this please?
Looks good and boot tested with kuli on s390.
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
> > From 3961f380bbe84a1036ddfc823039cbee31b44dcb Mon Sep 17 00:00:00 2001
> > From: =3D?utf-8?q?Fran=3DC3=3DA7ois=3D20Diakhat=3DC3=3DA9?=3D <fdiakh@g=
mail.com>
> > Date: Thu, 18 Mar 2010 14:48:20 +0100
> > Subject: virtio: console: Fix early_put_chars usage
> >=20
> > Currently early_put_chars is not used by virtio_console because it can =
only be used once a port has been found, at which point it's too late becau=
se it is no longer needed. This patch should fix it.
> > ---
> > drivers/char/virtio_console.c | 6 +++---
> > 1 files changed, 3 insertions(+), 3 deletions(-)
> >=20
> > diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_consol=
e.c
> > index f404ccf..691ba21 100644
> > --- a/drivers/char/virtio_console.c
> > +++ b/drivers/char/virtio_console.c
> > @@ -645,13 +645,13 @@ static int put_chars(u32 vtermno, const char *buf=
, int count)
> > {
> > struct port *port;
> > =20
> > + if (unlikely(early_put_chars))
> > + return early_put_chars(vtermno, buf, count);
> > +
> > port =3D find_port_by_vtermno(vtermno);
> > if (!port)
> > return 0;
> > =20
> > - if (unlikely(early_put_chars))
> > - return early_put_chars(vtermno, buf, count);
> > -
> > return send_buf(port, (void *)buf, count);
> > }
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Virtio_console usage of early printk
2010-03-22 17:45 Virtio_console usage of early printk François Diakhate
2010-03-23 2:16 ` Amit Shah
2010-03-23 7:42 ` Alexander Graf
@ 2010-03-23 8:55 ` Amit Shah
2010-03-23 14:58 ` François Diakhate
2 siblings, 1 reply; 6+ messages in thread
From: Amit Shah @ 2010-03-23 8:55 UTC (permalink / raw)
To: François Diakhate; +Cc: linuxppc-dev, virtualization
On (Mon) Mar 22 2010 [18:45:47], François Diakhate wrote:
> Hi all,
>
> As far as I can see, early_put_chars is not used by virtio_console
> because it checks whether there is a port available before using it.
> If I understand correctly, this makes it useless because once we have
> a port, we can use the regular virtio transport to output things to
> the console. Does the attached patch seem valid ? Feedback from s390
> and powerpc users who use this functionality would be appreciated.
>
> Thanks,
> François
> From 3961f380bbe84a1036ddfc823039cbee31b44dcb Mon Sep 17 00:00:00 2001
> From: =?utf-8?q?Fran=C3=A7ois=20Diakhat=C3=A9?= <fdiakh@gmail.com>
> Date: Thu, 18 Mar 2010 14:48:20 +0100
> Subject: virtio: console: Fix early_put_chars usage
>
> Currently early_put_chars is not used by virtio_console because it can only be used once a port has been found, at which point it's too late because it is no longer needed. This patch should fix it.
You'll have to give a signed-off-by for the patch.
Amit
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Virtio_console usage of early printk
2010-03-23 8:55 ` Amit Shah
@ 2010-03-23 14:58 ` François Diakhate
0 siblings, 0 replies; 6+ messages in thread
From: François Diakhate @ 2010-03-23 14:58 UTC (permalink / raw)
To: Amit Shah; +Cc: linuxppc-dev, virtualization
[-- Attachment #1: Type: text/plain, Size: 1295 bytes --]
On Tue, Mar 23, 2010 at 9:55 AM, Amit Shah <amit.shah@redhat.com> wrote:
> On (Mon) Mar 22 2010 [18:45:47], François Diakhate wrote:
>> Hi all,
>>
>> As far as I can see, early_put_chars is not used by virtio_console
>> because it checks whether there is a port available before using it.
>> If I understand correctly, this makes it useless because once we have
>> a port, we can use the regular virtio transport to output things to
>> the console. Does the attached patch seem valid ? Feedback from s390
>> and powerpc users who use this functionality would be appreciated.
>>
>> Thanks,
>> François
>
>> From 3961f380bbe84a1036ddfc823039cbee31b44dcb Mon Sep 17 00:00:00 2001
>> From: =?utf-8?q?Fran=C3=A7ois=20Diakhat=C3=A9?= <fdiakh@gmail.com>
>> Date: Thu, 18 Mar 2010 14:48:20 +0100
>> Subject: virtio: console: Fix early_put_chars usage
>>
>> Currently early_put_chars is not used by virtio_console because it can only be used once a port has been found, at which point it's too late because it is no longer needed. This patch should fix it.
>
> You'll have to give a signed-off-by for the patch.
>
> Amit
>
Sorry, I didn't have access to my computer this morning. Here is the
patch with a sign-off, if you still need it.
François
[-- Attachment #2: 0001-virtio-console-Fix-early_put_chars-usage.patch --]
[-- Type: text/x-patch, Size: 1265 bytes --]
From 209c553355a3ac04e68a9e9560cea90897588304 Mon Sep 17 00:00:00 2001
From: =?utf-8?q?Fran=C3=A7ois=20Diakhat=C3=A9?= <fdiakh@gmail.com>
Date: Thu, 18 Mar 2010 14:48:20 +0100
Subject: [PATCH] virtio: console: Fix early_put_chars usage
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Currently early_put_chars is not used by virtio_console because it can only be used once a port has been found, at which point it's too late because it is no longer needed. This patch should fix it.
Signed-off-by: François Diakhaté <fdiakh@gmail.com>
---
drivers/char/virtio_console.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
index f404ccf..691ba21 100644
--- a/drivers/char/virtio_console.c
+++ b/drivers/char/virtio_console.c
@@ -645,13 +645,13 @@ static int put_chars(u32 vtermno, const char *buf, int count)
{
struct port *port;
+ if (unlikely(early_put_chars))
+ return early_put_chars(vtermno, buf, count);
+
port = find_port_by_vtermno(vtermno);
if (!port)
return 0;
- if (unlikely(early_put_chars))
- return early_put_chars(vtermno, buf, count);
-
return send_buf(port, (void *)buf, count);
}
--
1.6.3.3
^ permalink raw reply related [flat|nested] 6+ messages in thread
end of thread, other threads:[~2010-03-23 15:18 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-03-22 17:45 Virtio_console usage of early printk François Diakhate
2010-03-23 2:16 ` Amit Shah
2010-03-23 8:17 ` Christian Borntraeger
2010-03-23 7:42 ` Alexander Graf
2010-03-23 8:55 ` Amit Shah
2010-03-23 14:58 ` François Diakhate
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).