From: Michael Buesch <mb@bu3sch•de>
To: Jeff Garzik <jeff@garzik•org>
Cc: bcm43xx-dev@lists•berlios.de,
Larry Finger <Larry.Finger@lwfinger•net>,
"John W. Linville" <linville@tuxdriver•com>,
netdev@vger•kernel.org
Subject: Re: Please pull 'upstream' branch of wireless-2.6
Date: Tue, 27 Jun 2006 17:25:25 +0200 [thread overview]
Message-ID: <200606271725.26037.mb@bu3sch.de> (raw)
In-Reply-To: <44A13C81.2050503@garzik.org>
On Tuesday 27 June 2006 16:11, Jeff Garzik wrote:
> Michael Buesch wrote:
> > On Tuesday 27 June 2006 04:27, Larry Finger wrote:
> >> Jeff Garzik wrote:
> >>> John W. Linville wrote:
> >>>> + assert(bcm->mac_suspended >= 0);
> >>>> + if (bcm->mac_suspended == 0) {
> >>>> + bcm43xx_power_saving_ctl_bits(bcm, -1, 1);
> >>>> + bcm43xx_write32(bcm, BCM43xx_MMIO_STATUS_BITFIELD,
> >>>> + bcm43xx_read32(bcm,
> >>>> BCM43xx_MMIO_STATUS_BITFIELD)
> >>>> + & ~BCM43xx_SBF_MAC_ENABLED);
> >>>> + bcm43xx_read32(bcm, BCM43xx_MMIO_GEN_IRQ_REASON); /* dummy
> >>>> read */
> >>>> + for (i = 100000; i; i--) {
> >>>> + tmp = bcm43xx_read32(bcm, BCM43xx_MMIO_GEN_IRQ_REASON);
> >>>> + if (tmp & BCM43xx_IRQ_READY)
> >>>> + goto out;
> >>>> + udelay(10);
> >>>> + }
> >>>> + printkl(KERN_ERR PFX "MAC suspend failed\n");
> >>>> }
> >>>
> >>> NAK this super-long delay... should be done in a workqueue, looks like?
> >>>
> >>> ACK everything else.
> >>>
> >> That delay was set to try to accommodate my interface when it refused to suspend the MAC, which
> >> resulted in transmit errors. That problem has since been cured by reworking the periodic work
> >> handlers - thus such a long delay should not be needed. The original spec from the clean-room group
> >> was a delay loop of 1000. I'm currently testing that value now. If it passes the test, would a for
> >> (i=1000; i; i--) be acceptable?
> >
> > Short: Don't touch it. Fullstop.
> >
> > Long: The delay will _never_ be exhausted. Actually the for-counter
> > is only there to not lock up the machine, if there is a Bug in the
> > driver. (__much__ easier debugging).
> > The loop will only iterate a few times, typically.
> > Actually, _if_ we want to change something, we should do this:
> >
> > for (i = 1000000; i; i--) {
> > ...
> > udelay(1);
> > }
> >
> > (max loop multiplied by 10, delay value divided by 10).
> > This will shorten the whole delay by a few usecs (up to 10).
> > I will send a patch for this, if it is desired.
> >
> > But lowering the loop counter value is NACKed by me,
> > because it simply does not make sense.
>
>
> My overriding concern was that this type of loop spins the CPU at 100%
> until the hardware condition is satisfied, which starves all other
> kernel work on that CPU, and is very unfriendly to power consumption
> (though I believe monitor/mwait/cpu_relax helps on x86).
Ok, I did a testrun. Here's the result:
[ 68.711243] bcm43xx_d80211: Chip ID 0x4306, rev 0x3
[ 68.712662] bcm43xx_d80211: Number of cores: 5
[ 68.714023] bcm43xx_d80211: Core 0: ID 0x800, rev 0x4, vendor 0x4243, enabled
[ 68.715536] bcm43xx_d80211: Core 1: ID 0x812, rev 0x5, vendor 0x4243, disabled
[ 68.717062] bcm43xx_d80211: Core 2: ID 0x80d, rev 0x2, vendor 0x4243, enabled
[ 68.718575] bcm43xx_d80211: Core 3: ID 0x807, rev 0x2, vendor 0x4243, disabled
[ 68.720089] bcm43xx_d80211: Core 4: ID 0x804, rev 0x9, vendor 0x4243, enabled
[ 68.724897] bcm43xx_d80211: PHY connected
[ 68.726154] bcm43xx_d80211: Detected PHY: Version: 2, Type 2, Revision 2
[ 68.727638] bcm43xx_d80211: Detected Radio: ID: 2205017f (Manuf: 17f Ver: 2050 Rev: 2)
[ 68.729232] bcm43xx_d80211: Radio turned off
[ 68.730512] bcm43xx_d80211: Radio turned off
[ 68.745153] wmaster0: Selected rate control algorithm 'simple'
[ 69.853876] bcm43xx_d80211: Virtual interface added (type: 0x00000002, ID: 7, MAC: 00:11:24:a0:de:8b)
[ 69.861872] bcm43xx_d80211: PHY connected
[ 70.000713] bcm43xx_d80211: Radio turned on
[ 70.190153] bcm43xx_d80211: Chip initialized
[ 70.192051] bcm43xx_d80211: DMA initialized
[ 70.193987] bcm43xx_d80211: 80211 cores initialized
[ 70.195550] bcm43xx_d80211: Keys cleared
[ 70.212565] wmaster0: Does not support passive scan, disabled
[ 70.252321] bcm43xx_d80211: mac_suspend() took 1 loops == 10 usec
[ 70.542160] NET: Registered protocol family 17
[ 70.692256] sta0: starting scan
[ 70.702234] HW CONFIG: channel=1 freq=2412 phymode=3
[ 70.762225] HW CONFIG: channel=2 freq=2417 phymode=3
[ 70.822227] HW CONFIG: channel=3 freq=2422 phymode=3
[ 70.882225] HW CONFIG: channel=4 freq=2427 phymode=3
[ 70.942225] HW CONFIG: channel=5 freq=2432 phymode=3
[ 71.002285] HW CONFIG: channel=6 freq=2437 phymode=3
[ 71.062229] HW CONFIG: channel=7 freq=2442 phymode=3
[ 71.122230] HW CONFIG: channel=8 freq=2447 phymode=3
[ 71.182239] HW CONFIG: channel=9 freq=2452 phymode=3
[ 71.242226] HW CONFIG: channel=10 freq=2457 phymode=3
[ 71.302226] HW CONFIG: channel=11 freq=2462 phymode=3
[ 71.512225] HW CONFIG: channel=1 freq=2412 phymode=3
[ 71.520066] sta0: scan completed
[ 71.523437] HW CONFIG: channel=6 freq=2437 phymode=3
[ 71.531600] sta0: Initial auth_alg=0
[ 71.531806] sta0: authenticate with AP 00:90:4c:60:04:00
[ 71.533279] sta0: RX authentication from 00:90:4c:60:04:00 (alg=0 transaction=2 status=0)
[ 71.533292] sta0: authenticated
[ 71.533301] sta0: associate with AP 00:90:4c:60:04:00
[ 71.536264] sta0: RX AssocResp from 00:90:4c:60:04:00 (capab=0x411 status=0 aid=1)
[ 71.536276] sta0: associated
[ 71.536352] wmaster0: Added STA 00:90:4c:60:04:00
[ 130.292251] bcm43xx_d80211: mac_suspend() took 1 loops == 10 usec
[ 190.322251] bcm43xx_d80211: mac_suspend() took 1 loops == 10 usec
[ 250.362252] bcm43xx_d80211: mac_suspend() took 0 loops == 0 usec
[ 310.392278] bcm43xx_d80211: mac_suspend() took 1 loops == 10 usec
So, I will submit a patch to lower the udelay(10) to udelay(1)
and we can close the discussion? ;)
--
Greetings Michael.
next prev parent reply other threads:[~2006-06-27 15:25 UTC|newest]
Thread overview: 96+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-06-26 21:25 Please pull 'upstream' branch of wireless-2.6 John W. Linville
2006-06-27 2:06 ` Jeff Garzik
2006-06-27 2:27 ` Larry Finger
2006-06-27 3:50 ` Jeff Garzik
2006-06-27 13:30 ` Michael Buesch
2006-06-27 14:11 ` Jeff Garzik
2006-06-27 14:34 ` Larry Finger
2006-06-27 14:36 ` Michael Buesch
2006-06-27 16:10 ` Jeff Garzik
2006-06-27 16:23 ` Michael Buesch
2006-06-27 15:25 ` Michael Buesch [this message]
2006-06-27 16:12 ` Jeff Garzik
2006-06-27 16:31 ` Michael Buesch
2006-06-27 19:33 ` John W. Linville
2006-06-27 19:47 ` Michael Buesch
2006-06-27 20:06 ` Larry Finger
2006-06-27 20:23 ` Michael Buesch
2006-06-27 20:37 ` Larry Finger
2006-06-28 14:34 ` Michael Buesch
2006-06-28 16:04 ` Larry Finger
2006-06-28 16:32 ` Michael Buesch
2006-06-28 17:32 ` Larry Finger
2006-06-28 18:02 ` Michael Buesch
2006-06-27 16:52 ` Joseph Jezak
2006-06-28 0:48 ` Please pull 'upstream' branch of wireless-2.6 (revised) John W. Linville
2006-06-28 17:36 ` Please pull 'zd1211rw' branch of wireless-2.6 (new driver) John W. Linville
2006-06-30 16:29 ` Jeff Garzik
2006-06-30 16:31 ` Please pull 'upstream' branch of wireless-2.6 (revised) Jeff Garzik
2006-06-30 19:27 ` John W. Linville
2006-07-01 15:00 ` John W. Linville
2006-07-05 18:25 ` Jeff Garzik
-- strict thread matches above, loose matches on Subject: below --
2007-05-29 18:30 Please pull 'upstream-fixes' branch of wireless-2.6 John W. Linville
2007-05-29 18:31 ` Please pull 'upstream' " John W. Linville
2007-05-30 14:03 ` Jeff Garzik
2007-05-08 17:39 John W. Linville
2007-05-09 22:54 ` Jeff Garzik
2007-05-07 17:51 John W. Linville
[not found] ` <20070507175121.GF5125-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org>
2007-05-07 21:15 ` Dan Williams
[not found] ` <1178572550.11805.7.camel-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
2007-05-07 22:51 ` John W. Linville
2007-05-08 8:49 ` Johannes Berg
2007-05-07 23:09 ` Jeff Garzik
[not found] ` <463FB1A0.3070608-o2qLIJkoznsdnm+yROfE0A@public.gmane.org>
2007-05-07 23:30 ` Michael Wu
2007-05-07 23:38 ` John W. Linville
2007-05-08 17:38 ` John W. Linville
2007-02-02 21:27 Please pull "upstream-fixes" " John W. Linville
2007-02-02 21:28 ` Please pull "upstream" " John W. Linville
2007-02-07 0:06 ` Please pull "upstream-fixes" " Jeff Garzik
[not found] ` <45C917EF.1060307-o2qLIJkoznsdnm+yROfE0A@public.gmane.org>
2007-02-07 21:11 ` Please pull "upstream" " John W. Linville
[not found] ` <20070207211117.GC6109-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org>
2007-02-09 20:13 ` John W. Linville
[not found] ` <20070209201306.GA2580-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org>
2007-02-09 21:12 ` Jeff Garzik
2007-01-18 15:48 Please pull 'upstream-fixes' " John W. Linville
2007-01-18 15:49 ` Please pull 'upstream' " John W. Linville
2007-01-19 3:10 ` Jeff Garzik
2007-01-19 8:42 ` John W. Linville
2007-01-23 5:36 ` Jeff Garzik
2007-01-03 2:41 Please pull 'upstream-fixes' " John W. Linville
2007-01-03 2:42 ` Please pull 'upstream' " John W. Linville
2007-01-18 12:16 ` John W. Linville
2006-12-21 3:03 Please pull 'upstream-fixes' " John W. Linville
2006-12-21 3:05 ` Please pull 'upstream' " John W. Linville
2006-12-26 21:39 ` Jeff Garzik
2006-12-28 0:10 ` John W. Linville
2007-01-03 2:04 ` John W. Linville
2006-12-12 0:21 John W. Linville
2006-12-06 1:42 John W. Linville
2006-12-07 10:03 ` Jeff Garzik
2006-11-15 1:29 Please pull 'upstream-fixes' " John W. Linville
2006-11-15 1:31 ` Please pull 'upstream' " John W. Linville
2006-11-28 19:13 ` John W. Linville
2006-11-08 4:58 Please pull 'upstream-fixes' " John W. Linville
2006-11-08 4:59 ` Please pull 'upstream' " John W. Linville
2006-11-08 19:48 ` John W. Linville
2006-11-14 15:29 ` Jeff Garzik
2006-10-17 21:34 Please pull 'upstream-fixes' " John W. Linville
2006-10-17 21:35 ` Please pull 'upstream' " John W. Linville
2006-10-21 18:22 ` Jeff Garzik
2006-09-11 23:58 Please pull 'upstream-fixes' " John W. Linville
2006-09-11 23:59 ` Please pull 'upstream' " John W. Linville
2006-09-12 15:43 ` Jeff Garzik
2006-09-12 19:49 ` Michael Buesch
2006-08-30 15:05 John W. Linville
2006-09-06 15:02 ` Jeff Garzik
2006-08-14 20:50 John W. Linville
2006-07-28 0:22 Please pull 'upstream-fixes' " John W. Linville
2006-07-28 0:23 ` Please pull 'upstream' " John W. Linville
2006-07-29 4:33 ` Jeff Garzik
2006-07-10 21:29 Please pull 'upstream-fixes' " John W. Linville
2006-07-10 21:31 ` Please pull 'upstream' " John W. Linville
2006-07-10 21:38 ` Michael Buesch
2006-07-10 21:58 ` Larry Finger
2006-07-19 17:51 ` Jeff Garzik
2006-06-15 20:03 John W. Linville
2006-06-20 8:46 ` Jeff Garzik
2006-06-05 21:53 Please pull 'upstream-fixes' " John W. Linville
2006-06-05 21:55 ` Please pull 'upstream' " John W. Linville
2006-06-08 19:48 ` Jeff Garzik
2006-05-22 19:18 Please pull 'upstream-fixes' " John W. Linville
2006-05-22 19:19 ` Please pull 'upstream' " John W. Linville
2006-05-24 4:35 ` Jeff Garzik
2006-05-24 12:42 ` John W. Linville
2006-05-17 19:34 Please pull 'upstream-fixes' " John W. Linville
2006-05-17 19:38 ` Please pull 'upstream' " John W. Linville
2006-05-17 21:23 ` Daniel Drake
2006-05-18 17:28 ` John W. Linville
2006-05-18 18:26 ` Daniel Drake
2006-05-06 1:06 Please pull upstream-fixes " John W. Linville
2006-05-06 1:09 ` Please pull upstream " John W. Linville
2006-04-24 19:40 Please pull 'upstream-fixes' " John W. Linville
2006-04-24 20:40 ` Please pull 'upstream' " John W. Linville
2006-04-25 0:33 ` Dan Williams
2006-04-25 11:30 ` Johannes Berg
2006-04-25 12:03 ` Dan Williams
2006-04-26 10:18 ` Jeff Garzik
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=200606271725.26037.mb@bu3sch.de \
--to=mb@bu3sch$(echo .)de \
--cc=Larry.Finger@lwfinger$(echo .)net \
--cc=bcm43xx-dev@lists$(echo .)berlios.de \
--cc=jeff@garzik$(echo .)org \
--cc=linville@tuxdriver$(echo .)com \
--cc=netdev@vger$(echo .)kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox