Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
Hello, Adrian. You wrote 16 мая 2013 г., 21:53:49: AC oh, when it drops to non-N mode, it likely did that because it failed AC to successfully transmit a BAR frame to the receiver. AC can you provide me the whole log? I'll go see what's going on. Attached. But I cannot say at which moment it drops to non-N mode now :( AC Ok. Update to the latest -HEAD again and retry. I've fixed debugging AC to be a little clearer (to me, at least.) Bad news: at this series of tests I didn't get N rates at all, and got station deassociation again. Twice. It seems, that it depends on HT/aggregation negotiation. Log attached. Log contains two series of tests: (1) Only debug was set to 0x9zeroes20 value (see sysctl call) Result: non-N rates, deassociation, automatic re-association in 30 seconds without my intervention. (2) Additional wlandebug +crypto +assoc +state was called. Result: the same. You could see borders of tests by sudo calls. Notebook was on AC power. -- // Black Lion AKA Lev Serebryakov l...@freebsd.org messages-r250706-tdma.log.xz Description: Binary data ___ freebsd-wireless@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-wireless To unsubscribe, send any mail to freebsd-wireless-unsubscr...@freebsd.org
Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
Hi! Yup, I've reproduced one of the TX hangs here. I'll post more details soon. Thanks! Adrian On 17 May 2013 01:45, Lev Serebryakov l...@freebsd.org wrote: Hello, Adrian. You wrote 16 мая 2013 г., 21:53:49: AC oh, when it drops to non-N mode, it likely did that because it failed AC to successfully transmit a BAR frame to the receiver. AC can you provide me the whole log? I'll go see what's going on. Attached. But I cannot say at which moment it drops to non-N mode now :( AC Ok. Update to the latest -HEAD again and retry. I've fixed debugging AC to be a little clearer (to me, at least.) Bad news: at this series of tests I didn't get N rates at all, and got station deassociation again. Twice. It seems, that it depends on HT/aggregation negotiation. Log attached. Log contains two series of tests: (1) Only debug was set to 0x9zeroes20 value (see sysctl call) Result: non-N rates, deassociation, automatic re-association in 30 seconds without my intervention. (2) Additional wlandebug +crypto +assoc +state was called. Result: the same. You could see borders of tests by sudo calls. Notebook was on AC power. -- // Black Lion AKA Lev Serebryakov l...@freebsd.org ___ freebsd-wireless@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-wireless To unsubscribe, send any mail to freebsd-wireless-unsubscr...@freebsd.org
Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
Hello, Adrian. You wrote 15 мая 2013 г., 22:34:26: AC I've just committed my first pass of improved power save and ps-poll AC support in AP mode to -HEAD. AC There's a bunch of corner cases I'm trying to solve but I'd rather AC this be in the driver now rather than later. AC So please, update to -HEAD and give this a whirl. Let me know if AC things are better, worse or the same. r250675, my ususal test (300 mbit AP - Client). Looks good. First of all, it gives me stable 100-150Mbit, so it could negotiate HT rates again without TCP :) And it sustain for 600 seconds in this mode, which was impossible before. There was several such messages: May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_suspend: tid=0xff8000d82230, bar_wait=0, bar_tx=0, called May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx_ready: c4:85:08:3f:9e:c2: tid=0xff8000d82230 (0), bar ready May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx: c4:85:08:3f:9e:c2: tid=0xff8000d82230, called May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx: c4:85:08:3f:9e:c2: tid=0xff8000d82230, new BAW left edge=676 May 16 12:57:31 gateway kernel: ath0: ath_bar_response: c4:85:08:3f:9e:c2: called; tap=0xff8000d7f3a0, atid=0xff8000d82230, txa_tid=0, atid-tid=0, status=0, attempts=1 May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_unsuspend: c4:85:08:3f:9e:c2: tid=0xff8000d82230, called But no full stalls. I have dev.ath.0.debug=0x90020 A lot of such messages in log when notebook is on battery (no such messages when it is on AC power): May 16 12:51:12 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:12 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:12 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:18 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:18 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:18 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:18 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:20 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:20 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:20 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:20 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:21 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:21 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:21 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:21 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:31 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:31 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:31 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:31 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:32 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:32 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:32 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:32 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:34 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 And sometimes it shows: May 16 13:05:52 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 13:05:52 gateway kernel: ath0: ath_node_recv_pspoll: c4:85:08:3f:9e:c2: leaking frame to TID 0 May 16 13:05:52 gateway kernel: ath0: ath_tx_leak_count_update: c4:85:08:3f:9e:c2: leak count = 1, psq=1, swq=6, MORE=1 May 16 13:05:52 gateway kernel: ath0: ath_node_recv_pspoll: c4:85:08:3f:9e:c2: leaking frame to TID 0 May 16 13:05:52 gateway kernel: ath0: ath_tx_leak_count_update: c4:85:08:3f:9e:c2: leak count = 1, psq=1, swq=8, MORE=1 May 16 13:05:52 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 It is when traffic goes 100-150Mbit/s, with drops to 75-85 every ~10th second. At second test, after ~300 seconds of test in battery mode it drops to non-N speeds and never recovered. -- // Black Lion AKA Lev Serebryakov l...@freebsd.org ___ freebsd-wireless@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-wireless To unsubscribe, send any mail to
Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
Hi! Those TIM messages and PSPOLL messages are expected. They're the power save state change.s Since your laptop on battery is likely trying ot save as much power as possible, this isn't surprising. can you give me the logging info when TX stalls? I'd like to see what the state of the TX queues are. Thanks! Adrian On 16 May 2013 02:09, Lev Serebryakov l...@freebsd.org wrote: Hello, Adrian. You wrote 15 мая 2013 г., 22:34:26: AC I've just committed my first pass of improved power save and ps-poll AC support in AP mode to -HEAD. AC There's a bunch of corner cases I'm trying to solve but I'd rather AC this be in the driver now rather than later. AC So please, update to -HEAD and give this a whirl. Let me know if AC things are better, worse or the same. r250675, my ususal test (300 mbit AP - Client). Looks good. First of all, it gives me stable 100-150Mbit, so it could negotiate HT rates again without TCP :) And it sustain for 600 seconds in this mode, which was impossible before. There was several such messages: May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_suspend: tid=0xff8000d82230, bar_wait=0, bar_tx=0, called May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx_ready: c4:85:08:3f:9e:c2: tid=0xff8000d82230 (0), bar ready May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx: c4:85:08:3f:9e:c2: tid=0xff8000d82230, called May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx: c4:85:08:3f:9e:c2: tid=0xff8000d82230, new BAW left edge=676 May 16 12:57:31 gateway kernel: ath0: ath_bar_response: c4:85:08:3f:9e:c2: called; tap=0xff8000d7f3a0, atid=0xff8000d82230, txa_tid=0, atid-tid=0, status=0, attempts=1 May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_unsuspend: c4:85:08:3f:9e:c2: tid=0xff8000d82230, called But no full stalls. I have dev.ath.0.debug=0x90020 A lot of such messages in log when notebook is on battery (no such messages when it is on AC power): May 16 12:51:12 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:12 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:12 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:18 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:18 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:18 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:18 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:20 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:20 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:20 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:20 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:21 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:21 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:21 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:21 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:31 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:31 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:31 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:31 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:32 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:32 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:32 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:32 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:34 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 And sometimes it shows: May 16 13:05:52 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 13:05:52 gateway kernel: ath0: ath_node_recv_pspoll: c4:85:08:3f:9e:c2: leaking frame to TID 0 May 16 13:05:52 gateway kernel: ath0: ath_tx_leak_count_update: c4:85:08:3f:9e:c2: leak count = 1, psq=1, swq=6, MORE=1 May 16 13:05:52 gateway kernel: ath0: ath_node_recv_pspoll: c4:85:08:3f:9e:c2: leaking frame to TID 0 May 16 13:05:52 gateway kernel: ath0: ath_tx_leak_count_update: c4:85:08:3f:9e:c2: leak count = 1, psq=1, swq=8, MORE=1 May 16 13:05:52 gateway kernel: ath0: ath_node_powersave:
Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
oh, when it drops to non-N mode, it likely did that because it failed to successfully transmit a BAR frame to the receiver. can you provide me the whole log? I'll go see what's going on. Thanks! adrian On 16 May 2013 02:09, Lev Serebryakov l...@freebsd.org wrote: Hello, Adrian. You wrote 15 мая 2013 г., 22:34:26: AC I've just committed my first pass of improved power save and ps-poll AC support in AP mode to -HEAD. AC There's a bunch of corner cases I'm trying to solve but I'd rather AC this be in the driver now rather than later. AC So please, update to -HEAD and give this a whirl. Let me know if AC things are better, worse or the same. r250675, my ususal test (300 mbit AP - Client). Looks good. First of all, it gives me stable 100-150Mbit, so it could negotiate HT rates again without TCP :) And it sustain for 600 seconds in this mode, which was impossible before. There was several such messages: May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_suspend: tid=0xff8000d82230, bar_wait=0, bar_tx=0, called May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx_ready: c4:85:08:3f:9e:c2: tid=0xff8000d82230 (0), bar ready May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx: c4:85:08:3f:9e:c2: tid=0xff8000d82230, called May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_tx: c4:85:08:3f:9e:c2: tid=0xff8000d82230, new BAW left edge=676 May 16 12:57:31 gateway kernel: ath0: ath_bar_response: c4:85:08:3f:9e:c2: called; tap=0xff8000d7f3a0, atid=0xff8000d82230, txa_tid=0, atid-tid=0, status=0, attempts=1 May 16 12:57:31 gateway kernel: ath0: ath_tx_tid_bar_unsuspend: c4:85:08:3f:9e:c2: tid=0xff8000d82230, called But no full stalls. I have dev.ath.0.debug=0x90020 A lot of such messages in log when notebook is on battery (no such messages when it is on AC power): May 16 12:51:12 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:12 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:12 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:18 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:18 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:18 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:18 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:20 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:20 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:20 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:20 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:21 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:21 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:21 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:21 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:31 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:31 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:31 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:31 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:32 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 May 16 12:51:32 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 12:51:32 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 May 16 12:51:32 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=0, an_pwrsave=0, disabling May 16 12:51:34 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=1 And sometimes it shows: May 16 13:05:52 gateway kernel: ath0: ath_node_set_tim: c4:85:08:3f:9e:c2: enable=1, enabling TIM May 16 13:05:52 gateway kernel: ath0: ath_node_recv_pspoll: c4:85:08:3f:9e:c2: leaking frame to TID 0 May 16 13:05:52 gateway kernel: ath0: ath_tx_leak_count_update: c4:85:08:3f:9e:c2: leak count = 1, psq=1, swq=6, MORE=1 May 16 13:05:52 gateway kernel: ath0: ath_node_recv_pspoll: c4:85:08:3f:9e:c2: leaking frame to TID 0 May 16 13:05:52 gateway kernel: ath0: ath_tx_leak_count_update: c4:85:08:3f:9e:c2: leak count = 1, psq=1, swq=8, MORE=1 May 16 13:05:52 gateway kernel: ath0: ath_node_powersave: c4:85:08:3f:9e:c2: enable=0 It is when traffic goes 100-150Mbit/s, with drops to 75-85 every ~10th second. At second
Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
On 16 May 2013 07:35, Lev Serebryakov l...@freebsd.org wrote: Hello, Adrian. You wrote 16 мая 2013 г., 18:20:50: AC oh, when it drops to non-N mode, it likely did that because it failed AC to successfully transmit a BAR frame to the receiver. AC can you provide me the whole log? I'll go see what's going on. Attached. But I cannot say at which moment it drops to non-N mode now :( Ok. Update to the latest -HEAD again and retry. I've fixed debugging to be a little clearer (to me, at least.) Adrian ___ freebsd-wireless@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-wireless To unsubscribe, send any mail to freebsd-wireless-unsubscr...@freebsd.org
Re: [rft] improved hostap powersave / ps-poll support is now in HEAD
Adrian Chadd adr...@freebsd.org writes: So please, update to -HEAD and give this a whirl. Let me know if things are better, worse or the same. The same here. ___ freebsd-wireless@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-wireless To unsubscribe, send any mail to freebsd-wireless-unsubscr...@freebsd.org