[Bug 548542] New: USB problems discovered when using the hso driver
http://bugzilla.novell.com/show_bug.cgi?id=548542 Summary: USB problems discovered when using the hso driver Classification: openSUSE Product: openSUSE 11.2 Version: RC 1 Platform: All OS/Version: openSUSE 11.2 Status: NEW Severity: Normal Priority: P5 - None Component: Kernel AssignedTo: kernel-maintainers@forge.provo.novell.com ReportedBy: nice@titanic.nyme.hu QAContact: qa@suse.de Found By: --- User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; hu-HU; rv:1.9.1.3) Gecko/20090909 SUSE/3.5.3-3.2 Firefox/3.5.3 My intention is to consider this bug as the "successor" of https://bugzilla.novell.com/show_bug.cgi?id=545982 which got too complex and contains some "false positives" too. So, to sum up, the symptom is the following: Since using the 2.6.31 kernel, my Option iCon 225 GSM modem became highly unreliable. To prove that the problem is in the kernel, I successfully started openSUSE 11.2 on the kernel of 11.1 (2.6.27) + hso 1.12. That combination worked extremely well, it was free of the problems described here. The problem is that sometimes (in fact quite often to make working impossible) the GSM modem seems to lose the connection to the hso driver. When this happens I see this in dmesg: [ 3428.891550] usb 3-2: reset full speed USB device using uhci_hcd and address 5 This mostly happens during connecting, when modem-manager instructed the modem to connect by sending 'AT_OWANCALL=1,1,1' to /dev/ttyHS0, and - according the LEDs on the modem - the connection succeeds virtually always, but the modem fails to inform the driver about the success, seemingly because of the error message above. I have to attempt 4-5, sometimes (seemingly depending on signal quality) even 8-10 times to connect so the modem finally be able to send back connection data to the driver. Sometimes this USB reset event also occurs during a successfuly initiated living session (which means I have to reconnect manually), but the driver, NetworkManager and modem-manager is not informed about it. Even the LEDs on the modem indicated a still living connection, I assume the connection gets lost only between the modem and the hso driver. When I load the hso kernel module with the option debug=1, I get some additional info in dmesg from the driver itself: [ 880.866932] [2062:put_rxbuf_data]: data to push to tty [ 880.868942] [2062:put_rxbuf_data]: data to push to tty [ 880.870941] [2062:put_rxbuf_data]: data to push to tty [ 885.087949] [1903:intr_callback]: Pending read interrupt on port 0 [ 885.087954] [ 885.090943] [2062:put_rxbuf_data]: data to push to tty [ 856.546940] [2062:put_rxbuf_data]: data to push to tty [ 856.548936] [2062:put_rxbuf_data]: data to push to tty [ 862.001950] [692:log_usb_status]: intr_callback: received USB status - endpoint not enabled (-2) [ 864.604075] usb 3-2: reset full speed USB device using uhci_hcd and address 10 [ 870.001950] [692:log_usb_status]: intr_callback: received USB status This problem didn't exist in 2.6.27+hso-1.12. That combination was almost alway able to successfully connect, and almost never lost connection. I don't know if the problem is caused by the usb code of the kernel, the driver itself or something else, but it seems to originate from inside the kernel. Reproducible: Always Steps to Reproduce: 1. Start openSUSE 11.2 with KDE 2. Insert an Option iCon 225 GSM modem 3. Create an appropriate GSM connection in knetworkmanager (or nm-applet) 4. Click on connect Actual Results: 4-5 attempts have to be made in order to connect to the GSM 3G network (and you will get disconnected in a few minutes anyway). Expected Results: It should work for firts attempt and work continously. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=548542#c1
--- Comment #1 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=548542#c2
--- Comment #2 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=548542#c3
--- Comment #3 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=548542#c4
--- Comment #4 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
User uwedr@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=548542#c5
Uwe Drechsel
http://bugzilla.novell.com/show_bug.cgi?id=548542
zhu rensheng
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c6
Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c7
--- Comment #7 from Tamás Németh
I had to go hunt down where to find that driver. Interestingly at the top of the post, it says: "If your kernel is 2.6.31 or higher do not use this driver! Instead use the driver supplied with the kernel."
I'm going to see what's needed to shape this up.
Thank you for takinf care for this bug. So, you may find some useful information at http://www.pharscape.org/forum/index.php?board=14.0 and especially http://www.pharscape.org/forum/index.php?topic=784.0 Read this latter topic carefully! You can find a patch there wich makes it possible to compile the hso-1.12 driver for the linux kernel 2.6.31. I've been using that for some week on openSUSE 11.2 and it performs absolutely perfectly except that becouse the killswitch feature is disabled, when the systems goes to suspend during and active 3G connection, the hso driver can not work anymore afterwards. The networking systems has to be stopped, the hso module unloaded and reloaded into the kernel, and the the networking system started again. You might be able to re-enabte that killswitch capability and find a way to include the hso-1.12 driver into the kernel of openSUSE 11.2 (and possibly SLED 11 SP 1 ;) in a smooth way. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c8
--- Comment #8 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c9
--- Comment #9 from Tamás Németh
Created an attachment (id=328776) --> (http://bugzilla.novell.com/attachment.cgi?id=328776) [details] Replacement hso.c
Can you replace your hso.c with this one and rebuild the module? This will be the first in a few tests since I don't actually have the hardware.
What? You've done an extremely good job! I replaced hso.ko with your version in the official openSUSE 11.2 kernel source, recompiled it, deleted to ot-of-the-kernel-tree driver, copied in the newly compiled one, ran depmod, and finally it works! Which version is it derived from? The kernel's own driver or hso-1.12? Anyway it seems to match the capabilities of my previous patched hso-1.12, at least it seems to be able to connect immediately but cannot survive s2ram (has to be unloaded and reloaded when suspended with an active connection). I'm going to test it in the next few days but it's already WAY better than 2.6.31's builtin module which is practically unusable. BTW, how can I compile only on kernel module? I tried with "make M=drivers/net/usb/hso" but didn't work for me. I hd to recompile the entire kernel, which took a few hours and grew to 5.2GB. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c10
--- Comment #10 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c11
--- Comment #11 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c12
--- Comment #12 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c13
--- Comment #13 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c14
--- Comment #14 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c15
--- Comment #15 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c16
Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c17
--- Comment #17 from Tamás Németh
Oops, comment #14 should have read version -jeffm5.
Ok, so now we have some test cases. I'm trying to accommodate the time difference since I assume you've just woken up or are about to go to bed very late. It's 3am my time, so I'm calling it a night.
Please test each of those attachments and let me know which one stops working correctly.
OK, all the six versions were able to be built, loaded into the kernel: milleniumfalcon:/usr/src/linux # dmesg | grep hso [ 119.468313] hso: drivers/net/usb/hso.c: 1.12-Option Option Wireless [ 119.469701] usbcore: registered new interface driver hso [ 121.767762] hso0: Disabled Privacy Extensions [ 1002.362215] hso: unloaded [ 1002.362238] usbcore: deregistering interface driver hso [ 1021.104514] hso: drivers/net/usb/hso.c: 1.12-Option-jeffm2 Option Wireless [ 1021.104913] usbcore: registered new interface driver hso [ 1023.489555] hso0: Disabled Privacy Extensions [ 1161.976085] hso: unloaded [ 1161.976109] usbcore: deregistering interface driver hso [ 1207.412607] hso: drivers/net/usb/hso.c: 1.12-Option-jeffm3 Option Wireless [ 1207.414452] hso0: Disabled Privacy Extensions [ 1207.416759] usbcore: registered new interface driver hso [ 1247.739209] hso 3-2:1.0: Not our interface [ 1250.302580] hso0: Disabled Privacy Extensions [ 1330.103085] hso: unloaded [ 1330.103109] usbcore: deregistering interface driver hso [ 1349.844991] hso: drivers/net/usb/hso.c: 1.12-Option-jeffm4 Option Wireless [ 1349.846203] usbcore: registered new interface driver hso [ 1352.236522] hso0: Disabled Privacy Extensions [ 1420.788376] hso: unloaded [ 1420.788401] usbcore: deregistering interface driver hso [ 1466.377470] hso: drivers/net/usb/hso.c: 1.12-Option-jeffm5 Option Wireless [ 1466.378720] hso0: Disabled Privacy Extensions [ 1466.382100] usbcore: registered new interface driver hso [ 1481.514229] hso 3-2:1.0: Not our interface [ 1484.078592] hso0: Disabled Privacy Extensions [ 1665.290309] hso: unloaded [ 1665.290336] usbcore: deregistering interface driver hso [ 1684.381580] hso: drivers/net/usb/hso.c: 1.12-Option-jeffm6 Option Wireless [ 1684.382802] usbcore: registered new interface driver hso [ 1686.770652] hso0: Disabled Privacy Extensions [ 1841.542296] hso 3-2:1.0: Not our interface [ 1844.108723] hso0: Disabled Privacy Extensions [ 1887.417291] hso 3-2:1.0: Not our interface [ 1889.982679] hso0: Disabled Privacy Extensions and - at the forst glance - every version seems to work as perfect as vanilla hso-1.12. However they (I only tried -jeffm6 so far) can't handle suspend/resume while the USB modem is plugged in. I finally figured out that the driver itselfy DOESN'T SUFFER ANY DAMAGE during suspend/resume, the only problem is that modem-manager somehow loses the modem, but the network interface (hso0) and the virtual serial interfaces (/dev/ttyHS*) remain fully intact and functional, I only have to kill modem-manager (which will be automatically restarted by NetworkManager) to regain the seemingly lost modem. After doing so I can work agani. So, to sum up, every version works well. Next time I will send you some modem-manager output extracted during suspend/resume. Thank you for the valuable work! -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c18
--- Comment #18 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c19
--- Comment #19 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c20
Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c21
--- Comment #21 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c22
--- Comment #22 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c23
--- Comment #23 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c
Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c24
Tamás Németh
Created an attachment (id=328830) --> (http://bugzilla.novell.com/attachment.cgi?id=328830) [details] Test hso.c #7
This one works, too. However, I was able to test it only in a well performing 3G environment ( http://www.rethink-wireless.com/2009/10/01/t-mobile-hungary-broadband-perfor... ). What if one uses it under circumstances like weak signal, overloaded network, frequent cell roaming, etc. or the combination of these? Shouldn't we be afraid of that this new driver would have problems in the mentioned difficult situations? Does it depend on the driver anyway?
Ok, hold onto that info. That will be a separate bug report and we'll involve people who know modem-manager more intimately.
I've already notified Tambet Ingo a few hours ago, but it's weekend now:) Many thanks for your efforts! -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c25
--- Comment #25 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c26
--- Comment #26 from Tamás Németh
Created an attachment (id=328832) --> (http://bugzilla.novell.com/attachment.cgi?id=328832) [details] Patched in-kernel driver
Ok, this is the in-kernel driver, patched to fix the reversed if statement. Can you try this one out?
This one does not work. It definitely has the issues described b this bugreport.
Obviously if your problems are made worse in areas with poor coverage, that would be good data.
I cannot try that. A long train or car journey would be necessary to do that.
If it usually works either way when you have good signal, then there's not a lot to work with.
I hope so. The hso 1.12-Option-jeffm7 version works almost perfectly except the suspend/resume problem, and sadly this one: https://bugzilla.novell.com/show_bug.cgi?id=545982#c7 Interestingly this latter symptom usually happens for the first (few?) occasions I plug in the modem, but not when using the system for days without a reboot. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c27
--- Comment #27 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c28
--- Comment #28 from Tamás Németh
Created an attachment (id=328833) --> (http://bugzilla.novell.com/attachment.cgi?id=328833) [details] Patched in-kernel driver #2
I tried it. It doesn't work either. It has all the problems hso-1.12 has and in addition it's practically unusable. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c29
--- Comment #29 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c30
Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c31
Tamás Németh
In response to comment #29, yep. That's expected. The Makefile in drivers/net/usb/ does the same thing as the Makefile chunk that I posted in comment #10.
I'm really at a loss to explain why test7 works and the version from comment #27 doesn't. They're virtually identical. Are you sure you used the right one?
I compared the two sources with diff, and realized that there is only a little difference between them, and yet despite this, the hso.c attached to c#27 does not work! I downloaded, compiled and installed it again very carefully, checked dmesg if I'm loading the correct version, and according to my experiences it doesn't work. I had many unsuccessful connection attempts and broken connections. If I can connect finally, the connection breaks so fast that I was even unable to write any comment here. Totally unusable. On the other side 1.12-jeffm7 is so stable for me that I haven't experienced any problem with it yet. Despite this I'd like to ask you to commit a slightly modified (fit for linux-2.6.31) hso-1.12 into the kernel, rather than a slightly modified (bugfixed) hso-1.2, because hso-1.12 it told (by Option) to have many fixes, and being a much more advanced driver than 1.2. If you finally decide what version to use where will you send it? Into vanilla 2.6.31.7/2.6.32 or Novell's own repositories? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c32
--- Comment #32 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c33
--- Comment #33 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c34
--- Comment #34 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c35
--- Comment #35 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c36
--- Comment #36 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c37
--- Comment #37 from Tamás Németh
Created an attachment (id=328856) --> (http://bugzilla.novell.com/attachment.cgi?id=328856) [details] Test hso.c #8
This version moves the wakeup and killing of the urb up in hso_stop_serial_device
Comparing this by diff to the one from c#27 doesn't show any semanical difference I can notice. And in fact this version also produces the usual in-kernel driver error. Just to remind you these errors are denoted by messages like "usb 3-2: reset full speed USB device using uhci_hcd and address 13" in dmesg. The bug may be hidden somewhere in the difference between this (or c#27's) version and c#23's version (which works). -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c39
--- Comment #39 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c40
--- Comment #40 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c41
--- Comment #41 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c42
--- Comment #42 from Tamás Németh
I'm really having a hard time believing that the c#23 version worked as expected and c#27/33 version doesn't work. I don't doubt that you've been testing. I just wonder if you've loaded the right version for some of your tests.
OK, I tested it again carefully for you. I downloaded hso.c from c#36 again, then compiled and loaded it. I double-checked everything during this process, and the results are the same. I'm getting loads of "reset full speed USB device" messages in dmesg, when using this driver. At my home, where the 3D signal is exceptionally good, the modem is sometimes able to recover from this problem, but now I'm at my working place, and I was even unable to do a single successful connection despite trying it many times. See the dmesg output showing the driver version: milleniumfalcon:~ # dmesg | egrep '(hso)|(reset full)' [ 6194.653345] hso: unloaded [ 6194.653371] usbcore: deregistering interface driver hso [ 6236.878437] hso: drivers/net/usb/hso.c: 1.2-test8 Option Wireless [ 6236.882215] hso0: Disabled Privacy Extensions [ 6236.884788] usbcore: registered new interface driver hso [ 6268.809107] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6311.059107] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6351.059084] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6390.809119] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6428.059108] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6449.809105] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6492.058110] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6549.059062] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6568.059112] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6612.058063] usb 3-2: reset full speed USB device using uhci_hcd and address 15 [ 6648.309116] usb 3-2: reset full speed USB device using uhci_hcd and address 15 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c43
--- Comment #43 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c44
Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c45
--- Comment #45 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c46
--- Comment #46 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c47
--- Comment #47 from Greg Kroah-Hartman
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c48
--- Comment #48 from Tamás Németh
I can not do a wholesale replacement of the driver, especially when the code is so different from the upstream kernel version.
And what about including it in the openSUSE kernels only?
If you could contact the upstream driver author, and ask them why their changes are not in the mainline kernel.org tree, that might be interesting to find out, especially as they should have pushed their changes a long time ago.
OK, I just did so. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c49
--- Comment #49 from Greg Kroah-Hartman
(In reply to comment #47)
I can not do a wholesale replacement of the driver, especially when the code is so different from the upstream kernel version.
And what about including it in the openSUSE kernels only?
Sorry, I can not do so, because then I would be forcing us to maintain this branch for "forever" :) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c50
Greg Kroah-Hartman
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c51
Tamás Németh
any response from upstream?
Subject: RE: Hso patch by Jeff Mahoney from novell
From: "Filip Aben"
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c52
--- Comment #52 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c53
--- Comment #53 from Greg Kroah-Hartman
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c54
--- Comment #54 from Tamás Németh
any response from upstream?
I saw this today: http://www.pharscape.org/article-1267446554.html Have you already received some update for 2.6.33.x? Can it be backported (or simply placed into) to 2.6.31? May I test it before an eventual backport? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c55
Greg Kroah-Hartman
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c56
--- Comment #56 from Tamás Németh
Can you try the latest FACTORY kernel to see if this resolves the issue for you now? That is based on 2.6.34-rc.
What is the exact URL of the repository holding the mentioned kernel? Is it http://download.opensuse.org/repositories/Kernel:/openSUSE-11.2/openSUSE_112... as said at https://bugzilla.novell.com/show_bug.cgi?id=590756#c1 ? I'd like to carry out he necessary test for both bugreports. Thank you in advance, and thank you for the kernel in general! (Its quality seems to be improving for me with each version nowadays!) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c57
--- Comment #57 from Jeff Mahoney
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c58
--- Comment #58 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c59
Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=548542
http://bugzilla.novell.com/show_bug.cgi?id=548542#c60
Greg Kroah-Hartman
participants (1)
-
bugzilla_noreply@novell.com