Back to Home / #uml / 2007 / 03 / Prev Day | Next Day
#uml IRC Logs for 2007-03-30

---Logopened Fri Mar 30 00:00:21 2007
01:23|-|ram [] has joined #uml
02:51|-|Urgleflogue [] has quit [Ping timeout: 480 seconds]
03:11|-|Urgleflogue [] has joined #uml
03:30|-|Urgleflogue [] has quit [Ping timeout: 480 seconds]
03:49|-|Urgleflogue [] has joined #uml
03:52|-|tyler [] has joined #uml
06:05|-|tyler [] has quit [Remote host closed the connection]
06:22|-|tyler [] has joined #uml
06:53|-|waldner [~waldner@] has joined #uml
06:53|-|waldner [~waldner@] has quit []
07:17|-|krau [~cktakahas@] has joined #uml
07:35|-|krau [~cktakahas@] has quit [Ping timeout: 480 seconds]
08:50|-|krau [~cktakahas@] has joined #uml
08:56|-|krau [~cktakahas@] has quit [Remote host closed the connection]
09:12|-|jdike [] has joined #uml
09:12<jdike>Hi guys
09:22<kokoko1>jdike, bonjour
09:36|-|krau [~cktakahas@] has joined #uml
09:51|-|tyler [] has quit [Ping timeout: 480 seconds]
09:57<peterz>jdike: does uml do something funny with fixmaps at the end of address space?
09:57<peterz>well, uhm,. I rewrote the whole mm -> vma lookup
09:58<peterz>and UML is the only thing that makes my kernel BUG
09:58<peterz>on exit_mm() there is still 1 nr_ptes left
09:58<jdike>i.e. UML BUGs?
09:59<jdike>there's nothing that UML can do that would make that happen
09:59<jdike>it's all just mmap and friends
09:59<jdike>nothing strange, that is
09:59<peterz>right, but its doing a _lot_ of them
09:59<peterz>I've got an mmap stress test prog
09:59<peterz>that works
09:59<peterz>full desktop bloat also works
10:00<peterz>just uml triggers this, so I was hoping it does something special
10:00<peterz>because I'm running out of ideas :-(
10:00<jdike>you're using a UML with highmem in use?
10:00<jdike>then there's no fixmap happening anyway
10:01<jdike>that's only a highmem thing, iirc
10:01<jdike>is it just the counter that got out of whack?
10:01<peterz>right, time to go write a pte scanner that looks for out of vma ptes
10:01<jdike>nr_ptes == 1 with an empty page table?
10:02<peterz>no idea, I wrote a pte scanner that goes up until the last vma vm_end address
10:02<peterz>and that didn't find any pte
10:02<peterz>so there might be one left between there and the end of address space
10:02<peterz>its just that when I put 0 or -1 in the end it goes into inf loops
10:02<jdike>there shouldn't be any ptes in non-VMA areas
10:02<peterz>but who knows, I might have messed up bad
10:03<peterz>RCU is a tricky thing
10:09|-|tyler [] has joined #uml
10:22<peterz>dOh, I should check up to TASK_SIZE of course
10:32<peterz>looks like there are indeed no ptes, so that would mean mm->nr_ptes is skewed
10:36|-|tyler [] has quit [Ping timeout: 480 seconds]
10:39|-|TBBle_ [~tbble@] has joined #uml
10:41|-|TBBle [~tbble@] has quit [Ping timeout: 480 seconds]
10:49|-|tyler [] has joined #uml
11:43|-|ram [] has quit [Read error: Operation timed out]
12:02|-|ram [] has joined #uml
12:45|-|tyler [] has quit [Ping timeout: 480 seconds]
13:02|-|tyler [] has joined #uml
13:36|-|tyler [] has quit [Remote host closed the connection]
13:37|-|baroni [] has quit [Ping timeout: 480 seconds]
13:53|-|tyler [] has joined #uml
13:59|-|fghj [~tkxue@agp.Stanford.EDU] has joined #uml
14:00|-|fghj [~tkxue@agp.Stanford.EDU] has quit []
14:02|-|tyler [] has quit [Ping timeout: 480 seconds]
14:07|-|ram [] has quit [Remote host closed the connection]
14:10|-|ram [] has joined #uml
14:12|-|baroni [] has joined #uml
14:16|-|tyler [] has joined #uml
14:45|-|tyler [] has quit [Ping timeout: 480 seconds]
15:02|-|tyler [] has joined #uml
15:30|-|tyler [] has quit [Ping timeout: 480 seconds]
15:50|-|tyler [] has joined #uml
16:33|-|hfb [] has joined #uml
16:46[~]kokoko1 pour some tea for jdike
16:51<jdike>as I walk back, tea cup in hand
16:56<kokoko1>I am done with mine :)
17:15|-|tyler [] has quit [Ping timeout: 480 seconds]
17:42|-|krau [~cktakahas@] has quit [Ping timeout: 480 seconds]
17:43|-|hfb [] has quit [Quit: Leaving]
17:54|-|fghj [~tkxue@agp.Stanford.EDU] has joined #uml
18:04|-|fghj [~tkxue@agp.Stanford.EDU] has quit [Quit: Leaving]
18:58<fo0bar>lovely, debian released a new 2.6.18 kernel today (-12, which will probably be the final before the etch release), which did not fix a bug in -11 that completely breaks compiling ARCH=um
18:59<fo0bar> <-- if anyone asks over the next 6 years :)
19:00<fo0bar>granted it's only with TT compiled, but ehh
20:00|-|ram [] has quit [Ping timeout: 480 seconds]
20:27|-|ram [] has joined #uml
21:01|-|ram [] has quit [Ping timeout: 480 seconds]
21:16|-|ram [] has joined #uml
21:45|-|jdike [] has quit [Quit: Leaving]
22:15|-|ram [] has quit [Ping timeout: 480 seconds]
22:27|-|ram [] has joined #uml
22:54|-|Netsplit <-> quits: tasaro, apic
22:56|-|Netsplit over, joins: tasaro, apic
22:59|-|VS_ChanLog [] has left #uml [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #uml
23:12|-|ram [] has quit [Ping timeout: 480 seconds]
23:41|-|caker [] has quit [Server closed connection]
23:41|-|caker [] has joined #uml
---Logclosed Sat Mar 31 00:00:49 2007