| === JanC is now known as Guest86327 | ||
| === JanC_ is now known as JanC | ||
| LocutusOfBorg | apw, we got probably the virtualbox kernel issue | 08:56 |
|---|---|---|
| LocutusOfBorg | >I haven't taken a look at the in-kernel version, but in RC1 the reason | 08:57 |
| LocutusOfBorg | >was that the arguments to out*() calls got swapped: the Linux kernel | 08:57 |
| LocutusOfBorg | >reverses them compared to everyone else I know of, and I failed to do | 08:57 |
| LocutusOfBorg | >that during clean-up to match kernel style. Will be committing soon. | 08:57 |
| apw | LocutusOfBorg, wonderful | 08:57 |
| apw | when will we have updated packages for that to take form ? | 08:58 |
| apw | from ? | 08:58 |
| LocutusOfBorg | I already asked for a patch | 09:10 |
| LocutusOfBorg | but if you have time, you can swap the parameters inside that outw call and see what happens | 09:11 |
| LocutusOfBorg | after re-enabling it | 09:11 |
| apw | sforshee, ^^ | 09:17 |
| sforshee | LocutusOfBorg: I grepped the vboxguest code in our kernel tree, I only found out* being used in one file (include/VBox/VBoxVideoGuest.h). The arguments are reversed, but best I can tell they aren't compiled in, looks like ASMOutU* is called instead, for example | 12:15 |
| sforshee | #ifdef VBOX_XPDM_MINIPORT | 12:15 |
| sforshee | VideoPortWritePortUchar((PUCHAR)Port, Value); | 12:15 |
| sforshee | #elif defined VBOX_GUESTR3XORGMOD | 12:15 |
| sforshee | outb(Port, Value); | 12:15 |
| sforshee | #else /** @todo make these explicit */ | 12:15 |
| sforshee | ASMOutU8(Port, Value); | 12:15 |
| sforshee | #endif | 12:15 |
| tjaalton | what's the problem with vbox? | 12:16 |
| tjaalton | sforshee: btw, when will next artful kernel be tagged? | 12:16 |
| sforshee | tjaalton: look at the code I pasted, the arguments to outb are reversed | 12:16 |
| sforshee | tjaalton: not sure, in theory the current artful-release kernel would be what we release with, so after release | 12:17 |
| sforshee | but if we have any critical bugs we may do another before then | 12:18 |
| tjaalton | there was some wifi bug that the new stable release fixed | 12:18 |
| sforshee | 4.13.5? | 12:19 |
| LocutusOfBorg | sforshee, I don't know honestly... I'll wait for the next vbox beta, and then see if at least vbox is fixed | 12:19 |
| LocutusOfBorg | in the meanwhile I hope somebody fixes the kernel too | 12:19 |
| sforshee | LocutusOfBorg: are you talking about the vboxvideo staging driver? | 12:20 |
| sforshee | I had to disable that in our kernel because it was causing problems | 12:20 |
| LocutusOfBorg | sforshee, yes, I'm talking about it | 12:21 |
| LocutusOfBorg | but the current vbox beta has the same trouble | 12:21 |
| LocutusOfBorg | because the kernel driver has been merged back into vbox code | 12:21 |
| tjaalton | sforshee: yep, 255432ecc6cbaf4 mac80211: fix deadlock in driver-managed RX BA session start | 12:21 |
| LocutusOfBorg | so, they "copied" the issue | 12:21 |
| LocutusOfBorg | and one issue discovered in vbox was that out* stuff | 12:22 |
| LocutusOfBorg | maybe this is not enough, I don't know | 12:22 |
| sforshee | LocutusOfBorg: okay. Thanks for the heads up, let us know when you have more information. But wrt to the mainline driver we won't have issues because it is disabled. | 12:23 |
| sforshee | maybe that is why we had problems | 12:23 |
| LocutusOfBorg | sure I know it is disabled | 12:23 |
| LocutusOfBorg | and the driver in 5.1.28 is working | 12:23 |
| LocutusOfBorg | fact is: the mainline driver is broken | 12:23 |
| LocutusOfBorg | vbox 5.2 beta2 was still ok | 12:24 |
| LocutusOfBorg | in 5.2 rc1 they merged the mainline into vbox again, and they broke vbox bundled driver too | 12:24 |
| LocutusOfBorg | they discovered such "swap" out function error, so maybe this is the problem in the kernel too | 12:24 |
| sforshee | tjaalton: how bad is that bug? Will it prevent users from installing or from updating after they install? | 12:24 |
| LocutusOfBorg | this is what I know, the next RC of vbox 5.2 will have a "fixed" driver | 12:24 |
| LocutusOfBorg | so I'll do some more testing and then see what the mainline kernel driver misses, and maybe we can re-enable it | 12:25 |
| tjaalton | sforshee: "Can't switch wireless connection between several WiFi APs" | 12:25 |
| tjaalton | so not too bad I guess | 12:25 |
| sforshee | tjaalton: yeah from that description at least sounds like something that could be fixed after release | 12:26 |
| gQuigs | from the linux-firmware changelog (artful) it should include the latest iwlwifi firmware - but the files differ | 16:10 |
| gQuigs | http://pastebin.ubuntu.com/25721005/ | 16:10 |
| gQuigs | my mistake! nvm | 16:28 |
| === cachio is now known as cachio_afk | ||
| === cachio_afk is now known as cachio | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!