| 00:06.24 | *** join/#utah brailsmt (~michael@12-254-226-160.client.attbi.com) |
| 00:11.44 | fugalh | hmmm. my motd has creeped again |
| 00:54.26 | Bradipo | Mecworks: you around? |
| 00:56.16 | *** join/#utah han_solo (~fugalh@andreas.fugal.net) |
| 00:56.57 | brailsmt | han_solo: you get that battery? |
| 00:57.07 | han_solo | a different one |
| 00:57.13 | han_solo | $46+shipping |
| 00:57.21 | han_solo | so $56. that was more comfortable |
| 00:57.31 | brailsmt | yeah |
| 00:57.55 | brailsmt | polin8 in #vim made an openbox theme based on midnight2.vim :) |
| 00:58.15 | han_solo | heh |
| 01:23.57 | *** join/#utah zollinger (~kevin@dsl093-224-185.slc1.dsl.speakeasy.net) |
| 01:24.10 | kryptik` | hello. |
| 01:24.31 | zollinger | Hola. |
| 01:36.24 | *** join/#utah levi_ (~levi@dpc6682175206.direcpc.com) |
| 01:39.58 | kryptik` | how often does harleypig stop by (#utah) |
| 01:40.01 | kryptik` | ? |
| 01:41.14 | *** join/#utah bbeattie (~bbeattie@cpe-66-1-180-69.ut.sprintbbd.net) |
| 01:47.58 | fugalh | ibot, seen harleypig? |
| 01:47.59 | | harleypig <~harleypig@208.186.100.102> was last seen on IRC in channel #utah, 48d 21h 11m 27s ago, saying: 'Anyone here familiar with Midgard and getting it setup?'. |
| 01:49.08 | kryptik` | thats cool |
| 01:49.12 | kryptik` | ibot, seen kryptik |
| 01:49.12 | | kryptik is currently on #utah (1d 17m 40s) |
| 01:49.18 | kryptik` | sweet |
| 02:32.07 | *** join/#utah levi__ (~levi@dpc6682175206.direcpc.com) |
| 03:15.03 | *** join/#utah kryptik` (Kryptik@208.187.113.123) |
| 03:15.09 | kryptik` | . |
| 03:22.18 | levi_ | , |
| 03:22.25 | fugalh | ; |
| 03:22.58 | levi_ | ! |
| 03:23.31 | fugalh | ~!@#$%^&*()_+-=[]{}|\":';?></., |
| 03:23.43 | levi_ | fugalh: Watch your language! |
| 03:24.14 | fugalh | :'/ |
| 03:24.30 | levi_ | Heh heh. |
| 03:25.53 | kryptik` | heh. |
| 03:26.28 | levi_ | Hah! |
| 03:26.34 | fugalh | ho! |
| 03:26.34 | kryptik` | Hah hah |
| 03:26.37 | kryptik` | ho ho |
| 03:26.43 | levi_ | Hee hee! |
| 03:26.48 | kryptik` | heh. |
| 03:26.54 | brailsmt | that perl? |
| 03:27.00 | kryptik` | lol |
| 03:27.02 | levi_ | kryptic: Ho. |
| 03:27.02 | fugalh | :) |
| 03:27.12 | kryptik` | levi_: I am not |
| 03:28.18 | levi_ | So, not a ho? D'oh! |
| 03:28.46 | kryptik` | heh. Mmm caffiene |
| 03:29.11 | levi_ | I had too many Dews today. |
| 03:29.30 | kryptik` | I think there is something wrong with me.. My stomach hurts like a mofo when I drink mt. dew, but I cant stop drinking it. |
| 03:29.56 | levi_ | Yes, that's called 'addiction'. :) |
| 03:30.25 | kryptik` | Good stuff eh? |
| 03:30.28 | kryptik` | hehe |
| 03:35.37 | brailsmt | mtn dew will shrink your family jewels |
| 03:36.23 | levi_ | Living leads to eventual death. |
| 03:36.54 | brailsmt | so lets accellerate it!!! |
| 03:36.59 | brailsmt | :) |
| 03:37.11 | levi_ | Live fast, die young! |
| 03:37.14 | brailsmt | of course mountain biking is hazardous to your health... |
| 03:37.19 | brailsmt | i guess i am doomed :) |
| 03:37.51 | levi_ | Who wants to be a decrepit old man anyway? |
| 03:39.14 | brailsmt | me |
| 03:39.29 | brailsmt | a mountain biking decrepit old man, that is :) |
| 03:39.52 | levi_ | Well, decrepit old men can't get on mountain bikes, thanks to the decrepit part |
| 03:40.13 | kryptik` | Man you guys have it down to a science. |
| 03:40.15 | kryptik` | :) |
| 03:40.45 | levi_ | it? |
| 03:41.07 | kryptik` | the whole mt dew thing. |
| 03:42.02 | levi_ | You know, dew only recently took on its 'extreme' image. |
| 03:42.09 | brailsmt | mt. dew contains yellow #5, which shrinks family jewels |
| 03:42.12 | levi_ | It used to be a hillbilly thing. |
| 03:42.31 | levi_ | About a billion other things contain yellow #5, too. |
| 03:42.33 | brailsmt | switch to code red, or vanilla coke |
| 03:42.47 | brailsmt | levi_: you are missing the point... ;) |
| 03:43.12 | levi_ | brailsmt: Cycling can cause impotence, too. |
| 03:43.31 | brailsmt | thats why i have the seat i have :p |
| 03:44.58 | levi_ | Hmm. The Yellow #5 thing is an urban legend, anyway. |
| 03:45.10 | brailsmt | i am not really concerned the shrinkage, just something i think is funny to bring up to mt. dew drinkers :) |
| 03:45.12 | levi_ | http://www.snopes.com/toxins/yellow.htm |
| 03:45.34 | brailsmt | you're no fun... |
| 03:45.39 | brailsmt | :) |
| 03:46.00 | levi_ | Perhaps not, but my manhood is intact! |
| 03:46.20 | brailsmt | by this discussion or mt. dew? |
| 03:46.49 | levi_ | Well, it hasn't been 'shrunk' by either. |
| 03:46.52 | brailsmt | i like code red |
| 03:46.59 | kryptik` | i like both |
| 03:47.09 | levi_ | It's not bad, but they don't stock it at work. |
| 03:47.10 | brailsmt | just not rocky mt. piss water |
| 03:48.08 | brailsmt | hey, i like to code vim syntax files too... *cough* |
| 03:58.26 | *** join/#utah kryptik` (Kryptik@208.187.113.123) |
| 05:02.13 | *** join/#utah fugalh (~fugalh@andreas.fugal.net) |
| 11:31.55 | *** join/#utah ChanServ (ChanServ@services.) |
| 11:31.55 | *** mode/#Utah [+o ChanServ] by calvino.freenode.net |
| 11:39.02 | *** join/#utah Bonez39 (~Bonez@206.71.67.22) |
| 15:37.27 | *** join/#utah fugalh (~fugalh@andreas.fugal.net) |
| 15:37.32 | fugalh | peer to peer is fascinating |
| 16:15.57 | fugalh | brailsmt: you there? can you get on blackboad? |
| 17:16.44 | *** join/#utah hAlcy0n (~halcyon@64.122.29.110) |
| 17:26.56 | *** join/#utah neybar (~jalance@dsl233.sisna.com) |
| 17:27.42 | Bradipo | What is blackboard? |
| 18:09.37 | *** join/#utah TimRiker (timr@rikers.org) |
| 18:09.37 | *** mode/#Utah [+o TimRiker] by ChanServ |
| 18:36.04 | *** join/#utah brailsmt (~michael@128.187.254.17) |
| 18:46.54 | *** join/#utah brailsmt (~michael@12-254-226-160.client.attbi.com) |
| 18:49.55 | brailsmt | i have a struct like this "struct data { int foo:26; int bar:6; };", and whenever I try 'memcpy(&my_struct, 0, sizeof(int))' i get a segfault... I don't know what i am doing wrong here... |
| 18:51.39 | brailsmt | nm |
| 18:51.56 | brailsmt | i thought i had already tried something i hadn't... |
| 18:52.23 | brailsmt | oops... that was supposed to be in #c++... :o |
| 19:04.04 | levi | Heh. |
| 19:04.45 | levi | Why not sizeof(struct data) ? |
| 19:06.34 | levi | You probably shouldn't assume that ints are 32 bits. |
| 19:10.45 | levi | Oh, and you want memset, not memcpy. |
| 19:11.02 | levi | It's segfaulting because you're trying to access memory at 0x0. |
| 19:13.12 | brailsmt | levi: i know that ints are 4 bytes... |
| 19:13.33 | brailsmt | and actually i have WORD_SIZE, #defined to be sozeof(uint) |
| 19:15.53 | brailsmt | this isn't going to be running on anything other that x86 |
| 19:15.58 | levi | In any case, I think it would be safe since the struct would fill to the full size of the int anyway. |
| 19:16.16 | levi | But that wasn't your problem, so it's moot. |
| 19:16.35 | brailsmt | it is a good point though... |
| 19:17.32 | brailsmt | the memset pointer is a good one though... |
| 19:17.59 | levi | Seems like if you're doing fixed-size fields in a struct (in portable code, at least) you should make sure that the type you're using is the right size for the job. |
| 19:18.54 | levi | Since typically you'd be doing that to save space, and if you compiled that on a 64 bit machine you'd have wasted 32 bits. |
| 19:19.18 | brailsmt | this is the memory management lab for cs 345... |
| 19:19.25 | levi | Unless you're trying to align on word boundaries or something. |
| 19:19.27 | brailsmt | it won't be running on anything other than x86 :) |
| 19:19.57 | brailsmt | levi: i am trying to split up 4 bytes in the various fields |
| 19:21.32 | levi | Non-Itanium x86, I assume. ;) |
| 19:22.00 | brailsmt | valid assumption :) |
| 19:22.16 | *** join/#utah Jayce^ (~jayce@65.204.214.20) |
| 19:22.17 | levi | 64 bit architectures will be hitting mainstream soon, though. |
| 19:22.36 | brailsmt | not in this semester in the CS labs... :) |
| 19:23.07 | brailsmt | which is the scope of this assignment :) |
| 19:23.26 | levi | I took CS143 3 or 4 times (don't ask). The architecture we used changed 3 times. But yeah, it shouldn't change mid-semester. |
| 19:23.58 | brailsmt | wow, 3-4 times... :o |
| 19:24.09 | Jayce^ | now we're all dying to ask.. |
| 19:24.38 | brailsmt | dangit, i used memset and now my unit tests fail... :( |
| 19:25.24 | levi | Well, it always started out really pathetically easy, so I skipped class a lot. Somewhere in the middle of the semester it must have started getting harder, 'cause I always got way off by then. |
| 19:26.11 | levi | The final time I took it, it was completely different and actually interesting and new stuff to me, so I payed attention through the whole semester and aced the class. |
| 19:26.47 | brailsmt | i think i got over 100% in that class... |
| 19:27.30 | brailsmt | of course this is my second go around in cs345 |
| 19:27.34 | levi | The time I passed it, I had one of the highest grades in the class. |
| 19:28.09 | brailsmt | third go around if you count the semester i dropped it after a month... |
| 19:30.32 | brailsmt | for some reason memset is not workng properly... |
| 19:31.01 | brailsmt | it copies the the low order bits into each bit field... |
| 19:31.40 | brailsmt | so memset(&foo, 0xdeadbeef, 4) => foo = f, f, eef, eef in the four fields... |
| 19:35.48 | brailsmt | stupid, c/c++ |
| 19:36.19 | levi | Well, are you trying to set it to 0 or set it to a default value? |
| 19:37.21 | brailsmt | i am trying to 1) set it to initial value which works... 2) set it to a value passed in, which works in my unit tests, but notwhen i use it in my program... |
| 19:37.55 | brailsmt | that is using memcpy... |
| 19:38.04 | brailsmt | with memset only the first instance works... |
| 19:38.12 | levi | You should be using memcpy if you have a value you want to copy in. |
| 19:38.30 | levi | You should be using memset if you want to clear it with 0s. |
| 19:39.45 | brailsmt | yeah, the problem i am having is that i have this function "void PTE::set_pte(uint val)" which has this as its body "memcpy(&data, &val, WORD_SIZE)" |
| 19:40.14 | brailsmt | it works properly in the unit test like this... "pte.set_pte(0xdeadbeef)" (among others) |
| 19:41.16 | brailsmt | but in my program "for(int idx = 0; idx < MAX_PAGE_TABLE_ENTRIES; idx++) { pte->set_pte(idx); }" <-- segfaults |
| 19:41.40 | brailsmt | it segfaults when idx == 1 |
| 19:42.13 | brailsmt | and i have absolutely no idea why it works in one case and not the other... |
| 19:45.10 | levi | Hmm. |
| 19:45.49 | levi | What's the memcpy call look like when it's segfaulting? |
| 19:46.11 | brailsmt | and when I put the code that segfaults into the unit test and call it exactly the same way, it works... |
| 19:46.30 | brailsmt | what do you mean? |
| 19:46.51 | levi | Like, what would strace say the last call was before it died? |
| 19:48.23 | brailsmt | well, loading the core file into gdb is how I know it segfaults on the memcpy... #0 of the stack trace is memcpy() from lib/libc.so.6 |
| 19:48.36 | levi | Ok, and what are the args to memcpy? |
| 19:49.19 | brailsmt | #1 is from PTE::set_pte(this = 0x1, val = 3221222804) at pte.cpp:28 |
| 19:50.03 | brailsmt | memcpy(&data, &val, WORD_SIZE) where WORD_SIZE is #define WORD_SIZE sizeof(uint) |
| 19:50.43 | levi | And what are &data, &val, data, and val? |
| 19:51.52 | brailsmt | gdb reports that val is 322122894, but i go up a frame and it is 1 |
| 19:52.28 | brailsmt | and it says data is 0x1?!?!... |
| 19:53.44 | brailsmt | i have pte=NULL in the loop... and i never create a new one... |
| 19:53.47 | levi | What are the addresses of data and val? Are they the same as the ones at the previous stack level? |
| 19:53.59 | brailsmt | voila |
| 19:54.27 | brailsmt | ok that only took 4 hours total time to find... |
| 19:54.40 | brailsmt | thx for your help |
| 19:55.19 | levi | Oh, hmm, 'this = 0x1' is a bad sign. :) That was due to pte=NULL, right? |
| 19:56.13 | brailsmt | yes |
| 19:56.26 | brailsmt | i was convinced the memcpy was causing the problem... |
| 19:56.34 | brailsmt | so i didn't look elsewhere... |
| 19:56.40 | levi | Ah, should have spotted that earlier. Not used to C++. |
| 19:56.43 | brailsmt | thats a bad habit of mine... |
| 19:57.00 | levi | Tsk. In C/C++, the problem is often not where you think it is! |
| 19:58.27 | brailsmt | i should have picked up on the this=0x1, but i didn't until i was trying to explain it to you... :| |
| 19:58.47 | brailsmt | ok, well no i know my page_table works :) |
| 19:59.07 | brailsmt | now to sublass it and make userpagetable and rootpagetable |
| 19:59.45 | brailsmt | levi: thx again for your help |
| 20:01.02 | brailsmt | well, UserPageTable was an easy class... :) |
| 20:01.28 | levi | You're welcome. :) |
| 20:02.41 | brailsmt | i wish C/C++ allowed '?' is identifiers... |
| 20:02.48 | brailsmt | s/is/in |
| 20:03.15 | brailsmt | makes things like "rpt->is_pinned?()" much more readable, IMO |
| 20:07.13 | brailsmt | builds a .cpp template based on the stuff it finds in the .h file :) |
| 20:07.23 | brailsmt | reduces typing immensely |
| 20:10.30 | levi | You could use the 'p' convention that lisps use. |
| 20:11.05 | levi | pinnedp |
| 20:11.58 | brailsmt | oh, yeah, i use the 'is' convention for that stuff, but ruby has me spoiled... |
| 21:26.16 | *** join/#utah Jayce^ (~jayce@65.204.214.20) [NETSPLIT VICTIM] |
| 21:26.16 | *** join/#utah levi (~levi@ash.lnxi.com) [NETSPLIT VICTIM] |
| 21:26.16 | *** join/#utah fozz (~fozz@12-255-4-85.client.attbi.com) [NETSPLIT VICTIM] |
| 21:26.16 | *** join/#utah Bradipo (~andyb@piedra.caldera.com) [NETSPLIT VICTIM] |
| 21:26.51 | *** join/#utah Bradipo (~andyb@piedra.caldera.com) [NETSPLIT VICTIM] |
| 21:26.51 | *** join/#utah fozz (~fozz@12-255-4-85.client.attbi.com) [NETSPLIT VICTIM] |
| 21:26.52 | *** join/#utah levi (~levi@ash.lnxi.com) |
| 21:26.52 | *** join/#utah Jayce^ (~jayce@65.204.214.20) [NETSPLIT VICTIM] |
| 21:26.52 | *** join/#utah Mecworks (marc@208.186.182.174) [NETSPLIT VICTIM] |
| 21:26.52 | *** join/#utah brailsmt (~michael@12-254-226-160.client.attbi.com) [NETSPLIT VICTIM] |
| 22:03.34 | Jayce^ | looks like ursula (leguin() is having troubles.. |
| 23:28.19 | *** join/#utah han_solo (~fugalh@andreas.fugal.net) |