> Based on my msleep patches, can you redo this one and your other ones I
> haven't applied yet?
Roger. I am back and resyncing. It'll take a few hours.
Regards
Oliver
---
This SF.Net email is sponsored by: SourceForge
On Tue, 18 May 2004, David Brownell wrote:
> Alan Stern wrote:
> >
> > I'm all in favor of something like this. Do you have any ideas on how we
> > should handle the case where the hub itself was disconnected but we don't
> > yet know that it's gone? We wouldn't want to log a KERN_ALERT messa
Look more closely at td_fill(), which morphs the dummy into td[0].
HeadP -> dummy
becomes
HeadP-> td[0] -> dummy'
where dummy' was previously td[0] and td[0] was previously dummy.
I think I get it. Actually dummy TD is fill by data of index=0,
so it morphs into td[0].
thanks & regards
yfw
_
Уважаемые господа!
Приглашаем Вас 24 мая посетить в однодневном семинаре на
тему : «Учет амортизируемого имущества»
Семинар рекомендуется главным бухгалтерам, налоговым консультантам, руководителям
пре6дприятия.
Программа семинара.
1. Понятие амортизируемого имущества. Различие
Alan Stern wrote:
On Wed, 5 May 2004, David Brownell wrote:
Hubs not disabling ports are severely broken; that code should probably
at least retry a bit before giving up, and giving up should syslog an
error at a severe level (maybe KERN_ALERT, telling user to unplug the
device). Until that probl
On Tue, May 18, 2004 at 09:20:40AM +0200, Oliver Neukum wrote:
> Am Dienstag, 18. Mai 2004 02:37 schrieb David Brownell:
> > >>this purges wait_ms() from ehci.
> > >
> > > David, look sane?
> >
> > I'd rather have a patch that covered msec_delay() too;
> > how about an updated version?
>
> Here
ChangeSet 1.1752, 2004/05/18 16:56:06-07:00, [EMAIL PROTECTED]
Some more misc wait_ms() conversions to use msleep()
drivers/macintosh/therm_pm72.c | 17 +
drivers/net/irda/stir4200.c |5 +++--
drivers/video/aty/radeon_base.c |2 +-
drivers/video/aty/ra
ChangeSet 1.1745, 2004/05/18 15:55:19-07:00, [EMAIL PROTECTED]
Remove libata's version of msleep()
drivers/scsi/libata-core.c | 17 -
1 files changed, 17 deletions(-)
diff -Nru a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
--- a/drivers/scsi/libata-core.c
ChangeSet 1.1747, 2004/05/18 15:57:20-07:00, [EMAIL PROTECTED]
USB: clean up usages of wait_ms() now that we have msleep()
drivers/usb/host/ehci-hcd.c |2 +-
drivers/usb/host/ehci-hub.c |2 +-
drivers/usb/input/aiptek.c |2 +-
drivers/usb/misc/usbtest.c
ChangeSet 1.1748, 2004/05/18 15:58:57-07:00, [EMAIL PROTECTED]
USB: remove ehci and ohci's private sleep function and use msleep() instead.
drivers/usb/host/ehci-hcd.c |4 ++--
drivers/usb/host/ehci.h |6 --
drivers/usb/host/ohci-hcd.c |4 ++--
drivers/usb/host/ohci-hub.c |
ChangeSet 1.1749, 2004/05/18 16:00:04-07:00, [EMAIL PROTECTED]
USB: remove wait_ms() from usb.h as it's no longer needed.
include/linux/usb.h | 11 ---
1 files changed, 11 deletions(-)
diff -Nru a/include/linux/usb.h b/include/linux/usb.h
--- a/include/linux/usb.h Tue May 18 1
ChangeSet 1.1751, 2004/05/18 16:51:33-07:00, [EMAIL PROTECTED]
Input: remove wait_ms() in place of using msleep()
drivers/input/gameport/ns558.c |5 +++--
drivers/input/gameport/vortex.c |3 ++-
drivers/input/joystick/adi.c|6 +++---
drivers/input/joystick/analog.c |6 +++--
The original message was received at Tue, 18 May 2004 20:21:01 -0400 (EDT)
from localhost [127.0.0.1]
- The following addresses had permanent fatal errors -
<[EMAIL PROTECTED]>
(reason: 550 Host unknown)
- Transcript of session follows -
550 5.1.2 <[EMAIL PROTECTED]>...
Hi,
Based on the previous USB changes creating a useful sleep function that
all drivers can use, Jeff Garzik suggested that I make it available to
all of the kernel. Doing this enabled me to remove a lot of local
driver functions that all did the same thing (and some of them did it
wrong, like th
ChangeSet 1.1750, 2004/05/18 16:00:39-07:00, [EMAIL PROTECTED]
I2C: change i2c_delay() to use msleep() instead.
drivers/i2c/busses/i2c-ali1535.c |5 +++--
drivers/i2c/busses/i2c-ali1563.c |5 +++--
drivers/i2c/busses/i2c-ali15x3.c |5 +++--
drivers/i2c/busses/i2c-amd756.c |7 ++
ChangeSet 1.1743, 2004/05/18 15:49:36-07:00, [EMAIL PROTECTED]
Add msleep function to the kernel core to prevent duplication.
include/linux/delay.h |2 ++
kernel/timer.c| 17 +
2 files changed, 19 insertions(+)
diff -Nru a/include/linux/delay.h b/include/linux/de
ChangeSet 1.1744, 2004/05/18 15:51:44-07:00, [EMAIL PROTECTED]
Delete block/carmel.c's version of msleep()
drivers/block/carmel.c |6 --
1 files changed, 6 deletions(-)
diff -Nru a/drivers/block/carmel.c b/drivers/block/carmel.c
--- a/drivers/block/carmel.cTue May 18 17:06:20 2004
On Tue, May 18, 2004 at 11:29:08AM -0400, Adam Kropelin wrote:
> This patch adds COMPATIBLE_IOCTLs for hiddev. Since all hiddev ioctl
> structures use explicitly sized types or int, COMPATIBLE_IOCTL should be
> sufficient for all archs. HIDIOCGNAME and HIDIOCGPHYS have been omitted
> because they p
On Tue, May 18, 2004 at 10:48:32PM +0200, Nemosoft Unv. wrote:
> Hi,
>
> On Tuesday 18 May 2004 18:19, you wrote:
> > On Sat, May 15, 2004 at 12:05:13AM +0200, Nemosoft Unv. wrote:
> > > -static int pwc_video_release(struct video_device *vfd)
> > > +static void pwc_video_release(struct video_devic
On Tue, May 18, 2004 at 09:46:54PM +0200, Oliver Neukum wrote:
>
> > + * msleep - sleep safely even with waitqueue interruptions
> > + * msecs: Time in milliseconds to sleep for
> > + */
> > +static inline void msleep(unsigned int msecs)
>
> Why do you make it inline? After a milisecond the cache
On Wed, 5 May 2004, David Brownell wrote:
> Hubs not disabling ports are severely broken; that code should probably
> at least retry a bit before giving up, and giving up should syslog an
> error at a severe level (maybe KERN_ALERT, telling user to unplug the
> device). Until that problem is reso
> + * msleep - sleep safely even with waitqueue interruptions
> + * msecs: Time in milliseconds to sleep for
> + */
> +static inline void msleep(unsigned int msecs)
Why do you make it inline? After a milisecond the cache is cold,
so we should shrink the code.
Regards
Oliv
Here's another revision for usb_kill_urb(). This one adds new fields for
both the reject flag and the usage counter, so there's no possibility of
usb_get_urb() breaking things.
Although Oliver claimed the usage counter could be a single bit, that's
not correct. It needs to be able to count at le
On Tue, May 18, 2004 at 02:41:34PM -0400, Jeff Garzik wrote:
> Greg KH wrote:
> >Doh. Ok, how about this?
> >
> >And yes, I'll kill the usb specific version, and the other uses of
> >msleep in the kernel tree when I send it off to Linus.
>
> Looks OK except for...
>
>
> >= include/linux/wai
Greg KH wrote:
Doh. Ok, how about this?
And yes, I'll kill the usb specific version, and the other uses of
msleep in the kernel tree when I send it off to Linus.
Looks OK except for...
= include/linux/wait.h 1.10 vs edited =
--- 1.10/include/linux/wait.h Sat May 15 04:50:38 2004
+++ edi
On Tue, May 18, 2004 at 01:23:11PM -0400, Jeff Garzik wrote:
>
> It's mostly pointless to inline delay functions :) That was one of the
> things I liked about the implementation you just committed to upstream,
> that it wasn't inlined.
>
> Stick it kernel/timer.c maybe?
Doh. Ok, how about th
Greg KH wrote:
On Tue, May 18, 2004 at 12:30:06PM -0400, Jeff Garzik wrote:
Your attached patch needs to be generic, just like the recent
msecs_to_jiffies() stuff...
I call this function "msleep" in my code, though your implementation is
better (msleep is still a better name, though :))
Ok, doe
On Tue, May 18, 2004 at 12:30:06PM -0400, Jeff Garzik wrote:
>
> Your attached patch needs to be generic, just like the recent
> msecs_to_jiffies() stuff...
>
> I call this function "msleep" in my code, though your implementation is
> better (msleep is still a better name, though :))
Ok, does
VIRUS ALERT
Our virus checker found
virus: Worm/NetSky.P
banned filename: message.scr
in your email to the following recipient:
-> [EMAIL PROTECTED]
Delivery of the email was stopped!
Please check your system for viruses,
or ask your system administrator to do so.
For your reference, he
On Sat, May 15, 2004 at 12:05:13AM +0200, Nemosoft Unv. wrote:
> -static int pwc_video_release(struct video_device *vfd)
> +static void pwc_video_release(struct video_device *vfd)
> {
> - Trace(TRACE_OPEN, "pwc_video_release() called. Now what?\n");
> + Trace(TRACE_OPEN, "pwc_video_release
fengwei yin wrote:
kernel 2.4.x
Doesn't matter, all Linuxes use roughly the same code for that.
Sorry,I can't find where HeadP is set to td[0].
Look more closely at td_fill(), which morphs the dummy into td[0].
HeadP -> dummy
becomes
HeadP-> td[0] -> dummy'
where dummy' was previously td[0] and
From: David Brownell <[EMAIL PROTECTED]>
To: fengwei yin <[EMAIL PROTECTED]>
CC: [EMAIL PROTECTED], [EMAIL PROTECTED]
Subject: Re: [linux-usb-devel] get hub port status on big endian board
Date: Tue, 18 May 2004 06:26:03 -0700
fengwei yin wrote:
I need guidance About the TD queue processing.
Let me
On Tue, May 18, 2004 at 10:47:10AM +0200, Vojtech Pavlik wrote:
> On Mon, May 17, 2004 at 09:58:55AM -0400, Adam Kropelin wrote:
> > On Mon, May 17, 2004 at 04:03:20PM +0400, Sergey Vlasov wrote:
> > > On Sun, May 16, 2004 at 02:58:03PM -0700, Greg KH wrote:
> > > > Ick, does hiddev really need hid
On Tue, 18 May 2004, Roman wrote:
> That is, TDs one transaction can settle down in different Queues or
> there is a rule of their arrangement?
All the TDs for one transaction must belong to the same queue.
> Please, describe what TDs it is necessary to send the controller and
> in what order, b
fengwei yin wrote:
I need guidance About the TD queue processing.
Let me take usb_set_address as example.
Before TD queue processing,the TD queue as following:
TailP
|
|
I haven't been following this thread, so I may have misunderstood
everything. That said...
> > Shouldn't "msecs" be "unsigned"? and then something like this
If msecs is the amount of time you are going to delay for, it maybe
better to leave it signed.
> Right. A negative delay would have inter
On Mon, May 17, 2004 at 09:58:55AM -0400, Adam Kropelin wrote:
> On Mon, May 17, 2004 at 04:03:20PM +0400, Sergey Vlasov wrote:
> > On Sun, May 16, 2004 at 02:58:03PM -0700, Greg KH wrote:
> > > Ick, does hiddev really need hid.h? If so, then we need to move it to
> > > include/linux/.
> >
> > Ho
I need guidance About the TD queue processing.
Let me take usb_set_address as example.
Before TD queue processing,the TD queue as following:
TailP
|
|
V
Hi,
I overlooked some.
Sorry
Oliver
You can import this changeset into BK by piping this whole message to:
'| bk receive [path to repository]' or apply the patch as usual.
===
[EMAIL PROTECTED], 2004-05-18
Am Dienstag, 18. Mai 2004 02:43 schrieb David Brownell:
> Oliver Neukum wrote:
>
> > +void usb_uninterruptible_sleep_ms(int msecs)
> > +{
> > + long timeout = msecs_to_jiffies(msecs);
>
> while in
>
>static inline unsigned long msecs_to_jiffies(unsigned int m)
>
> Shouldn't "msecs" be "u
Am Dienstag, 18. Mai 2004 02:37 schrieb David Brownell:
> >>this purges wait_ms() from ehci.
> >
> > David, look sane?
>
> I'd rather have a patch that covered msec_delay() too;
> how about an updated version?
Here you are. It's an additional change set.
Regards
Oliver
41 matches
Mail list logo
