Bug 15382 (om26er)
Summary: | ath5k: phy0: failed to wakeup the MAC Chip | ||
---|---|---|---|
Product: | Drivers | Reporter: | Omer Akram (om26er) |
Component: | network-wireless | Assignee: | drivers_network-wireless (drivers_network-wireless) |
Status: | CLOSED INSUFFICIENT_DATA | ||
Severity: | normal | CC: | adundovi, amichai2, dispiste, harviecz, linville, me, mickflemm |
Priority: | P1 | ||
Hardware: | All | ||
OS: | Linux | ||
Kernel Version: | 2.6.32.8 | Subsystem: | |
Regression: | No | Bisected commit-id: |
Description
Omer Akram
2010-02-24 10:46:40 UTC
And also I never Hibernate/suspend this happens to me when I am surfing internet/downloading something Exactly the same problem here. Moreover, about half an hour after getting this error, my laptop hard freezes, which is really really bad. I've got the same problem (also acer aspire) with 2.6.33: https://bugzilla.kernel.org/show_bug.cgi?id=ath5k-wakeup (i've created another bug, because at the beggining it seemed to be destroying EEPROM...) ath5k guys, ping? One thought, aspire has problems all over the map. Maybe this is related to the speed of the device, e.g. we aren't meeting timing requirements for booting the mac chip due to slow cpu, different races being exposed, etc. I may be able to replicate this in kvm. Omer: BTW are you using cpufreq? i do and now i realized that CPU (BTW: Mobile AMD Sempron(tm) Processor 3000+) may be scaled down from 1800MHz to 800MHz during suspend/wakeup - but probably not during reboot. Bob: does frequency scaling matters? anyway 800MHz CPU should handle 802.11abg... there are much slower machines (routers,APs) runing the same hardware at even lower frequencies... My netbook fails to suspend all the time so the problem I face happens randomly like something is downloading at highspeed or the internet has been up for several hours. and yes cpu scaling is on by default in ubuntu. Bob: do you think that frequency scaling can affect something? Reported downstream at https://bugs.launchpad.net/ubuntu/+source/linux/+bug/432353, perhaps some of the info in the comments there might be useful in diagnosing/fixing this... Bob: there's some allmost year old patch: https://patchwork.kernel.org/patch/38550/ * Don't put chip to full sleep because there are problems during wakeup. Instead hold MAC/Baseband on warm reset state via a new function ath5k_hw_on_hold. * Durring attach preserve pcicfg bits when enabling pci core sw retry fix. * Minor cleanups Signed-off-by: Nick Kossifidis <mickflemm@gmail.com> Such approach seems reasonable to me... until ve will be able to wakeup properly... Oh... that patch was applied to all the kernels which i have unsuccessfully tested, so it's probably useless (if reverting it can't change anything)... Any progress on this one? This sounds like the ASPM problem... try turning on CONFIG_PCIEASPM if you don't have it on already. Closing due to lack of response -- please reopen if/when requested information becomes available... Same problem here, and it's really frustrating (so i'm willing to help as much as i can to resolve this issue) :-) Summary: Wireless unexpectedly dies (didn't find exact reason, probably because of havy bandwidth load). Power laptop off and on brings wireless up again (ath5k module reload, restart and so doesn't help). I tested it on Fedora 12 and newest ArchLinux (2.6.32 -> 2.6.35). I don't remember that same was happening on older kernels (before 2.6.32). It doesn't depend on WL encryption (WEP/WPA) and network client (wicd, NetworkManager). Can you please tell us what card you use ? When ath5k loads you'll see something like this on dmesg... ath5k phy1: Atheros AR2413 chip found (MAC: 0x78, PHY: 0x45) ...this way we 'll know at least if we are talking about a pci or a pci-e card etc. Then if it's a pci-e card please be sure that your kernel is compiled with CONFIG_PCIEASPM as Bob said to rule out an already known problem related to pci-e cards. Device: $ dmesg | grep "ath5k phy" ath5k phy0: Atheros AR2425 chip found (MAC: 0xe2, PHY: 0x70) It's pci-e ( http://www.atheros.com/pt/AR5007EG.htm ). $ zcat /proc/config.gz | grep -i aspm # CONFIG_PCIEASPM is not set I'll try to recompile kernel with this option set so I'll be back with results (but probably not today). Tnx. I've recomiled kernel few hours ago with CONFIG_PCIEASPM=y, tested with heavy download, tried suspend/hibernate and (for now) wireless works without problems. So, I think that this issue is closed (for my bug report) :-) Tnx once again! |