s3erios_gmail.com (Andriy Voskoboinyk)
User

Projects

User does not belong to any projects.

User Details

User Since
May 26 2015, 5:01 PM (121 w, 3 d)

Recent Activity

Wed, Aug 30

s3erios_gmail.com added inline comments to D12157: [net80211] add A-MSDU hardware decap sequence number range checking This is part one of a multi-part series aimed at supporting hardware decap'ed A-MSDU frames. Hardware decap'ed A-MSDU frames have a variety of unfun behaviours: * for now, they....
Wed, Aug 30, 12:06 AM

Aug 22 2017

s3erios_gmail.com updated the diff for D11785: upgt: add some overflow checks.

Add some checks against actlen.

Aug 22 2017, 9:47 PM

Aug 21 2017

s3erios_gmail.com updated the diff for D11785: upgt: add some overflow checks.

Add memory wrap check.

Aug 21 2017, 8:11 AM
s3erios_gmail.com updated the diff for D11785: upgt: add some overflow checks.

Turn some KASSERT's into checks + add more size checks.

Aug 21 2017, 2:43 AM

Aug 20 2017

s3erios_gmail.com added inline comments to D11786: uath: fix varible types, add checks for descriptor / command header structure fields..
Aug 20 2017, 10:43 AM
s3erios_gmail.com updated the diff for D11786: uath: fix varible types, add checks for descriptor / command header structure fields..
Aug 20 2017, 10:38 AM

Aug 13 2017

s3erios_gmail.com added inline comments to D11785: upgt: add some overflow checks.
Aug 13 2017, 8:22 PM
s3erios_gmail.com updated the diff for D11786: uath: fix varible types, add checks for descriptor / command header structure fields..
  • Add more 'framelen' variable checks.
  • Check hdr->len before using it to calculate dlen.
Aug 13 2017, 7:51 PM
s3erios_gmail.com added inline comments to D11786: uath: fix varible types, add checks for descriptor / command header structure fields..
Aug 13 2017, 7:16 PM

Aug 7 2017

s3erios_gmail.com updated the diff for D11786: uath: fix varible types, add checks for descriptor / command header structure fields..

Address some possible buffer over-reads.

Aug 7 2017, 12:04 AM

Jul 30 2017

s3erios_gmail.com created D11786: uath: fix varible types, add checks for descriptor / command header structure fields..
Jul 30 2017, 10:49 PM
s3erios_gmail.com created D11785: upgt: add some overflow checks.
Jul 30 2017, 10:37 PM

Jul 23 2017

s3erios_gmail.com created D11705: rtwn_usb: add support for fragmented Rx.
Jul 23 2017, 5:44 PM

May 21 2017

s3erios_gmail.com added a comment to D10728: iwn: (partially) rewrite A-MPDU Tx path.

I can't test iwn(4) with 4965AGN - firmware throws FH_ERROR's during testing with and/or without this patch when HT rates are in use (A-MPDU / HT40 / SGI status does not matter).
Scanning is not reliable too - sometimes AP is shown on (almost) all 5GHz channels, while it is present on 44th channel only -> this results in NMI_INTERRUPT_WDG error during next scan.

May 21 2017, 5:32 PM

May 16 2017

s3erios_gmail.com added inline comments to D10728: iwn: (partially) rewrite A-MPDU Tx path.
May 16 2017, 7:16 AM
s3erios_gmail.com created D10753: net80211: clear i_seq for A-MPDU frames.
May 16 2017, 6:34 AM

May 15 2017

s3erios_gmail.com added a comment to D10728: iwn: (partially) rewrite A-MPDU Tx path.

node / mbuf NULL checks now are omitted

not really omitted - just considered as 'correct' - and ignored (see SCD_QUERY status comment + workarounds cannot cover all possible cases).

May 15 2017, 12:34 AM
s3erios_gmail.com created D10728: iwn: (partially) rewrite A-MPDU Tx path.
May 15 2017, 12:29 AM

Apr 9 2017

s3erios_gmail.com accepted D10328: [net80211] refactor out "add slot" and "purge slot" for A-MPDU This is in preparation for A-MSDU decap in A-MPDU support. * refactor out the code to purge a single reorder slot into ampdu_rx_purge_slot(). * refactor out the code to add a frame....
Apr 9 2017, 8:27 AM

Apr 3 2017

s3erios_gmail.com accepted D10240: [net80211] refactor out code into ampdu_dispatch_slot(). This is again in preparation for A-MSDU in A-MPDU offload support, where A-MSDUs are decapped into standalone 802.11/802.3 MPDUs but with the same seqno/crypto IV...
Apr 3 2017, 6:39 AM

Apr 1 2017

s3erios_gmail.com accepted D10208: [net80211] refactor the AMPDU RX window stuff!.
Apr 1 2017, 5:57 PM

Mar 26 2017

s3erios_gmail.com created D10146: iwm: fix build without IWM_DEBUG.
Mar 26 2017, 5:50 PM

Mar 22 2017

s3erios_gmail.com created D10106: iwn: deduplicate some parts of iwn_tx_data() and iwn_tx_data_raw().
Mar 22 2017, 10:02 PM

Mar 19 2017

s3erios_gmail.com added a comment to D9984: net80211: add a timer to flush fast-frames queues.

Yeah, this creates unneeded overhead for permanent data flow (and not so noticeable at the end, since it called only once); I will remove it.
BTW, there is another problem: with default ieee80211_ffagemax there is almost no chances to get an aggregate; typically, the queue is flushed earlier.

I was seeing two-frame AMSDUs during UDP TX tests. I'll update my test devices and re-test all of this and get back to you.

Mar 19 2017, 10:29 PM
s3erios_gmail.com added a comment to D9984: net80211: add a timer to flush fast-frames queues.

Yeah, this creates unneeded overhead for permanent data flow (and not so noticeable at the end, since it called only once); I will remove it.
BTW, there is another problem: with default ieee80211_ffagemax there is almost no chances to get an aggregate; typically, the queue is flushed earlier.

Mar 19 2017, 9:50 PM

Mar 13 2017

s3erios_gmail.com added a comment to D9984: net80211: add a timer to flush fast-frames queues.

The whole M_AGE thing is a bit "special". For fast frames I bet yeah, it's just the right age for the first frame. For the say power save queue, it's supposed to be the age of the frame as a delta from the previous one. If it isn't doing that then it's a bug.

Mar 13 2017, 8:11 AM
s3erios_gmail.com added a comment to D9984: net80211: add a timer to flush fast-frames queues.

Hm, do you think it'd be worth to do something like "ieee80211_tx_end() / ieee80211_rx_end()" as a kind of driver-driven "I've finished TX/RX loop, please do any deferred/interesting work now" in drivers?

I would rather there be something that was generic versus the special cased SUPERG stuff. (And that way we don't need to ifdef it for superg.)

Mar 13 2017, 8:00 AM
s3erios_gmail.com accepted D9986: [net80211] begin migrating WME parameters to be per-vap..

Probably wme_state[4] (in struct ieee80211_wme_state, ieee80211_proto.h) should be removed too?

Mar 13 2017, 7:41 AM
s3erios_gmail.com added a comment to D9984: net80211: add a timer to flush fast-frames queues.

IMHO, M_AGE_*() is useless for both powersave & stage queues - the first frame always has AGE set to some default value + it is always 0 for subsequent frames; how is it supposed to work?

Mar 13 2017, 12:25 AM
s3erios_gmail.com created D9984: net80211: add a timer to flush fast-frames queues.
Mar 13 2017, 12:17 AM

Mar 3 2017

s3erios_gmail.com updated the diff for D9871: net80211: fix ieee80211_htrateset setup, return EINVAL for an unsupported ucast/mcast/mgmt HT rate.

Init htrateset a bit later

Mar 3 2017, 12:25 AM
s3erios_gmail.com retitled D9871: net80211: fix ieee80211_htrateset setup, return EINVAL for an unsupported ucast/mcast/mgmt HT rate from net80211: fix ieee80211_rateset setup, return EINVAL for an unsupported ucast/mcast/mgmt HT rate to net80211: fix ieee80211_htrateset setup, return EINVAL for an unsupported ucast/mcast/mgmt HT rate.
Mar 3 2017, 12:23 AM
s3erios_gmail.com retitled D9871: net80211: fix ieee80211_htrateset setup, return EINVAL for an unsupported ucast/mcast/mgmt HT rate from to net80211: fix ieee80211_rateset setup, return EINVAL for an unsupported ucast/mcast/mgmt HT rate.
Mar 3 2017, 12:22 AM

Feb 26 2017

s3erios_gmail.com abandoned D5055: iwn: remove internal taskqueue.

It was additionaly reused for 'rftoggle_task' (and may be reused few more times in future).

Feb 26 2017, 7:52 PM
s3erios_gmail.com retitled D9811: net80211 drivers: fix rate setup for EAPOL frames, obtain Tx parameters directly from the node from to net80211 drivers: fix rate setup for EAPOL frames, obtain Tx parameters directly from the node.
Feb 26 2017, 7:48 PM
s3erios_gmail.com retitled D9806: net80211: handle VHT nodes in ieee80211_node_setuptxparms() from to net80211: handle VHT nodes in ieee80211_node_setuptxparms().
Feb 26 2017, 3:51 PM

Feb 25 2017

s3erios_gmail.com retitled D9797: iwn: some initialization / RF switch state change fixes from to iwn: some initialization / RF switch state change fixes.
Feb 25 2017, 12:27 AM

Feb 12 2017

s3erios_gmail.com updated the diff for D9559: snd_hda(4): reserve one more sequence number (14) for headphones redirection.

s/has/have (twice)

Feb 12 2017, 7:36 PM
s3erios_gmail.com retitled D9559: snd_hda(4): reserve one more sequence number (14) for headphones redirection from to snd_hda(4): reserve one more sequence number (14) for headphones redirection.
Feb 12 2017, 7:14 PM

Jan 30 2017

s3erios_gmail.com added a comment to D9369: net80211: cleanup ic_flags / iv_flags.

Hmm, so what to do with IEEE80211_F_COMP / IEEE80211_F_TXPOW_FIXED (AUTO/OFF) ?

Jan 30 2017, 10:22 PM

Jan 29 2017

s3erios_gmail.com accepted D9357: [net80211] address seqno allocation for group addressed frames.
Jan 29 2017, 10:27 AM

Jan 28 2017

s3erios_gmail.com retitled D9369: net80211: cleanup ic_flags / iv_flags from to net80211: cleanup ic_flags / iv_flags.
Jan 28 2017, 8:22 PM
s3erios_gmail.com added inline comments to D9357: [net80211] address seqno allocation for group addressed frames.
Jan 28 2017, 3:32 PM
s3erios_gmail.com added a comment to D9195: [net80211] initial "driver does sequence number handling" This implements: * optionally deferring sequence number allocation into the driver * don't grab/release the net80211 TX lock if the driver does sequence number allocation * write a....

Other parts look fine.

Jan 28 2017, 3:11 PM

Jan 27 2017

s3erios_gmail.com added inline comments to D9357: [net80211] address seqno allocation for group addressed frames.
Jan 27 2017, 8:30 PM

Jan 23 2017

s3erios_gmail.com retitled D9308: Garbage collect IFT_IEEE80211 from to Garbage collect IFT_IEEE80211.
Jan 23 2017, 10:30 PM

Jan 21 2017

s3erios_gmail.com retitled D9277: ath: adapt LDPC support checks from to ath: adapt LDPC support checks.
Jan 21 2017, 4:25 PM
s3erios_gmail.com added a comment to D9268: net80211: allow to configure LDPC support.

Ideally nodes without NODE_HT / NODE_VHT set won't even do HT/VHT bits. If we have that happening we should fix it. :-)

Jan 21 2017, 1:02 PM
s3erios_gmail.com updated the diff for D9268: net80211: allow to configure LDPC support.

Fix few typos

Jan 21 2017, 11:02 AM
s3erios_gmail.com added a comment to D9268: net80211: allow to configure LDPC support.

I'm unsure about IEEE80211_NODE_* flags setup in ieee80211_ht.c - if user will play enough with ifconfig(8) parameters some nodes may be created without the flag but with ni->ni_htcap and vap->iv_flags_ht bits set.

Jan 21 2017, 12:06 AM
s3erios_gmail.com retitled D9268: net80211: allow to configure LDPC support from to net80211: allow to configure LDPC support.
Jan 21 2017, 12:00 AM

Jan 18 2017

s3erios_gmail.com added inline comments to D9195: [net80211] initial "driver does sequence number handling" This implements: * optionally deferring sequence number allocation into the driver * don't grab/release the net80211 TX lock if the driver does sequence number allocation * write a....
Jan 18 2017, 11:25 PM

Jan 12 2017

s3erios_gmail.com updated the diff for D9021: rtwn: export more stats to net80211.
  • Remove IEEE80211_RX_F_DECRYPTED flag setup.
  • Call ieee80211_add_rx_params() in rtwn_rx_common() (and drop __DECONST); ieee80211_input_mimo() will check mbuf tag presence anyway.
  • Fix possible NULL pointer dereference (after ieee80211_find_rxnode() call).
  • Do not try to define R92C_RX_DRVINFO_SZ_DEF in if_rtwnvar.h; it is always included before r92c_reg.h.
  • Fix rate mask in r12a_rx_radiotap_flags().
Jan 12 2017, 12:17 AM

Jan 2 2017

s3erios_gmail.com retitled D9021: rtwn: export more stats to net80211 from to rtwn: export more stats to net80211..
Jan 2 2017, 2:35 PM

Dec 25 2016

s3erios_gmail.com added a comment to D8903: rtwn: add radar detection support for RTL8821AU (disabled by default).

Yes, this is USB NIC.

Dec 25 2016, 10:16 PM
s3erios_gmail.com retitled D8903: rtwn: add radar detection support for RTL8821AU (disabled by default) from to rtwn: add radar detection support for RTL8821AU (disabled by default).
Dec 25 2016, 6:53 PM

Dec 21 2016

s3erios_gmail.com accepted D8804: iwn: DMA fixes, and race condition fix.
Dec 21 2016, 9:22 PM

Dec 15 2016

s3erios_gmail.com added inline comments to D8804: iwn: DMA fixes, and race condition fix.
Dec 15 2016, 9:29 PM
s3erios_gmail.com added inline comments to D8804: iwn: DMA fixes, and race condition fix.
Dec 15 2016, 7:12 AM

Dec 10 2016

s3erios_gmail.com accepted D8717: Fix iwn(4) on ARM: perform BUS_DMASYNC_PREREAD when initializing RX buffer.
Dec 10 2016, 8:47 AM

Dec 4 2016

s3erios_gmail.com retitled D8706: rsu: add hardware crypto support (WEP, TKIP and CCMP) from to rsu: add hardware crypto support (WEP, TKIP and CCMP).
Dec 4 2016, 1:30 AM
s3erios_gmail.com accepted D8704: Switch run(4) to use a bitmap for debug printf()s.
Dec 4 2016, 12:17 AM

Nov 27 2016

s3erios_gmail.com abandoned D5151: rsu: cancel scan (via ieee80211_cancel_scan()) when it fails.

Committed as a part of rS309210

Nov 27 2016, 1:16 PM

Nov 21 2016

s3erios_gmail.com added inline comments to D8365: [net80211] add initial support for HT upgrade of IBSS nodes. * add ability for IBSS nodes to be upgraded when it sees HT IEs * add HT IEs to probe requests/responses and beacons * send a probe request after bringing up a node so we can....
Nov 21 2016, 8:47 PM
s3erios_gmail.com added inline comments to D8365: [net80211] add initial support for HT upgrade of IBSS nodes. * add ability for IBSS nodes to be upgraded when it sees HT IEs * add HT IEs to probe requests/responses and beacons * send a probe request after bringing up a node so we can....
Nov 21 2016, 8:35 PM
s3erios_gmail.com accepted D8365: [net80211] add initial support for HT upgrade of IBSS nodes. * add ability for IBSS nodes to be upgraded when it sees HT IEs * add HT IEs to probe requests/responses and beacons * send a probe request after bringing up a node so we can....
Nov 21 2016, 8:08 PM

Nov 18 2016

s3erios_gmail.com added a comment to D8571: Fix tx power (diff) calculation for rtl8812au/rtl8821au.

It does not use these fields (1T1R); that's why it went unnoticed.

Nov 18 2016, 9:05 PM
s3erios_gmail.com accepted D8571: Fix tx power (diff) calculation for rtl8812au/rtl8821au.

Yes, this is off-by-one bug. Thanks!

Nov 18 2016, 8:57 PM

Nov 17 2016

s3erios_gmail.com accepted D8553: [net80211] migrate IEEE80211_C bits out into _ieee80211.h for applications to use..
Nov 17 2016, 10:45 PM
s3erios_gmail.com accepted D8539: [net80211] Don't check bgscanidle setting in net80211 for full-offload scan..

Ok, thanks for the description; I've just misread the code.

Nov 17 2016, 8:44 PM
s3erios_gmail.com added a comment to D8539: [net80211] Don't check bgscanidle setting in net80211 for full-offload scan..

the NIC driver might look at this setting and pass it on to the firmware in some way if possible.

Nov 17 2016, 12:31 AM

Nov 15 2016

s3erios_gmail.com accepted D8533: [net80211] handle hardware encryption offload in the receive path.
Nov 15 2016, 11:56 PM

Nov 11 2016

s3erios_gmail.com accepted D8499: iwm(4): Don't panic if firmware load fails.
Nov 11 2016, 7:38 PM

Oct 29 2016

s3erios_gmail.com added inline comments to D8365: [net80211] add initial support for HT upgrade of IBSS nodes. * add ability for IBSS nodes to be upgraded when it sees HT IEs * add HT IEs to probe requests/responses and beacons * send a probe request after bringing up a node so we can....
Oct 29 2016, 8:51 AM
s3erios_gmail.com added inline comments to D8364: [net80211] extend/add crypto key flags for more key / payload offload. * Add flags to say "don't mmic/iv insert", etc - ath10k firmware and other boards want it. * Move the demic routine into ieee80211_crypto.c and add support for being told....
Oct 29 2016, 8:40 AM
s3erios_gmail.com added inline comments to D8364: [net80211] extend/add crypto key flags for more key / payload offload. * Add flags to say "don't mmic/iv insert", etc - ath10k firmware and other boards want it. * Move the demic routine into ieee80211_crypto.c and add support for being told....
Oct 29 2016, 8:34 AM
s3erios_gmail.com added inline comments to D8364: [net80211] extend/add crypto key flags for more key / payload offload. * Add flags to say "don't mmic/iv insert", etc - ath10k firmware and other boards want it. * Move the demic routine into ieee80211_crypto.c and add support for being told....
Oct 29 2016, 8:32 AM

Oct 18 2016

s3erios_gmail.com added a comment to D8262: [net80211] begin teaching the scan layer about full offload..

eg, things like "cancel scan", "is scan finished" methods to query the driver may be useful here.

Oct 18 2016, 9:53 PM

Oct 16 2016

s3erios_gmail.com accepted D8262: [net80211] begin teaching the scan layer about full offload..

So, firmware-based scanning will not be implemented as an 'ieee80211_scan_fw' module?

Oct 16 2016, 9:09 PM

Oct 11 2016

s3erios_gmail.com added a comment to D8207: [net80211/wifi bits]: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter.

Thanks for this!

Oct 11 2016, 7:37 PM

Oct 9 2016

s3erios_gmail.com added a comment to D8207: [net80211/wifi bits]: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter.

P.S. is there any need to call memcpy(3) in ieee80211_???_rx_params() ? As I can see, it can return a pointer to ieee80211_rx_params structure instead (NULL if operation fails)

Oct 9 2016, 9:59 PM
s3erios_gmail.com retitled D8207: [net80211/wifi bits]: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter from to [net80211/wifi bits]: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter.
Oct 9 2016, 9:54 PM
s3erios_gmail.com closed D8037: ieee80211_ratectl*: switch to reusable KPI.

Committed in rS306591

Oct 9 2016, 9:45 PM

Sep 26 2016

s3erios_gmail.com updated D8037: ieee80211_ratectl*: switch to reusable KPI.
Sep 26 2016, 6:17 PM
s3erios_gmail.com added a dependent revision for D7996: net80211: add one-vap version of ieee80211_iterate_nodes(): D8037: ieee80211_ratectl*: switch to reusable KPI.
Sep 26 2016, 6:10 PM
s3erios_gmail.com added a dependency for D8037: ieee80211_ratectl*: switch to reusable KPI: D7996: net80211: add one-vap version of ieee80211_iterate_nodes().
Sep 26 2016, 6:10 PM
s3erios_gmail.com retitled D8037: ieee80211_ratectl*: switch to reusable KPI from to ieee80211_ratectl*: switch to reusable KPI.
Sep 26 2016, 6:09 PM

Sep 21 2016

s3erios_gmail.com retitled D7996: net80211: add one-vap version of ieee80211_iterate_nodes() from to net80211: add one-vap version of ieee80211_iterate_nodes().
Sep 21 2016, 10:52 PM
s3erios_gmail.com accepted D7994: [iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames.

Looks good to me.

Sep 21 2016, 8:32 PM
s3erios_gmail.com added a comment to D7979: net80211: do not cancel scan when packet is sent.

Hm, don't we have a general scan timeout that should prevent a scan from hanging forever?

Sep 21 2016, 8:04 PM

Sep 20 2016

s3erios_gmail.com abandoned D3827: ieee80211_radiotap.9: reflect changes from D3826.

Committed in rS306049.

Sep 20 2016, 6:58 PM
s3erios_gmail.com abandoned D3826: ath(4), rum(4), run(4): remove TSF from radiotap headers in the transmit path.

Committed in rS306049.

Sep 20 2016, 6:57 PM
s3erios_gmail.com retitled D7979: net80211: do not cancel scan when packet is sent from to net80211: do not cancel scan when packet is sent.
Sep 20 2016, 2:35 PM
s3erios_gmail.com accepted D7959: [net80211] check IBSS SSID before adding nodes discovered by beacons/probe requests..

Works fine when 'hidessid' is not set (however, 'hidessid' bug is not related to this change).

Sep 20 2016, 11:08 AM
s3erios_gmail.com added inline comments to D7959: [net80211] check IBSS SSID before adding nodes discovered by beacons/probe requests..
Sep 20 2016, 10:22 AM

Sep 15 2016

s3erios_gmail.com added a comment to D7901: net80211: allow to override SWCRYPT/SWMIC bits in the driver.

Probably, yes (e.g., 88EU supports 64 MAC ID slots but has only 32 key slots; 12AU/21AU - 128 MAC ID slots, 64 key slots)

Sep 15 2016, 10:11 PM
s3erios_gmail.com updated the test plan for D7901: net80211: allow to override SWCRYPT/SWMIC bits in the driver.
Sep 15 2016, 9:56 PM
s3erios_gmail.com updated D7901: net80211: allow to override SWCRYPT/SWMIC bits in the driver.
Sep 15 2016, 9:47 PM
s3erios_gmail.com retitled D7901: net80211: allow to override SWCRYPT/SWMIC bits in the driver from to net80211: allow to override SWCRYPT/SWMIC bits in the driver.
Sep 15 2016, 9:45 PM

Sep 13 2016

s3erios_gmail.com retitled D7887: net80211: improve error checking in ieee80211_parse_{wpa,rsn} from to net80211: improve error checking in ieee80211_parse_{wpa,rsn}.
Sep 13 2016, 9:38 PM