irclog2html for uclibc on 2002.09.04

00:16:13sjhillandersee: works! thx
13:49:27sjhillandersee: hi
13:49:29anderseesjhill: morning
13:49:39sjhillsorry to bother you
13:49:43anderseesjhill: The idea is, each malloc can provide its own memalign
13:50:11anderseesjhill: Then valloc is shared, and built on top of whatever memalign is currently present
13:50:17anderseesjhill: no problem
13:50:29anderseesjhill: That is why I get paid the big bucks.  ;-)
13:50:41sjhillheh
13:51:25anderseesjhill: make sense?
13:51:35sjhilloh yes
13:52:09anderseeIdeally, we would have just one memalign, but I suspect that will be hard...
13:52:27sjhill'malloc-930716 doesn't have 'memalign'....should i not be using that?
13:52:40sjhillwhat i mean is, currently my Config uses 93701
13:53:23anderseemalloc-930716 does have a memalign -- its hiding in the Attic
13:53:42anderseeI killed it, in the hope of getting a common one
13:54:10sjhilli shall reincarnate it then
13:54:18sjhillspeaks incantation
13:54:56anderseeremember to stir the goat blood only 3 times...
13:55:05sjhillahh, right... :)
14:01:06sjhillon a very positive note, it looks like all my LTP patches were approved
14:01:42sjhillnext we can look at uclibc vs. glibc performance somewhat
14:05:29anderseeVery nice
14:05:55anderseeI saw your msg (yesterday?) re almost working sans valloc
14:06:05sjhillyup
14:06:32anderseeI'm very interested in seeing how we do, and seeing what (if anything) fails
14:14:39sjhillthis is a fricking nightmare
14:15:09sjhilli've brought back memalign.c free.c and things still won't compile...all kinds of undefined heap variables
14:17:06sjhillandersee: i'm wait out of my element on this
14:17:15sjhilli have no fricking clue what i'm doing
14:19:55sjhilli don't have time for this, i'm just going to set my Config to use 'malloc' instead
14:20:02sjhillscre 930716
14:20:04sjhillw
14:22:16anderseesjhill: why free?
14:22:38anderseewill look at it
14:25:08sjhilli appreciate it, memory allocation algorithms scare me...a lot
14:31:55sjhilldisappears for a bit
14:31:56sjhillbbl
14:43:26sjhillandersee: ok, i haven't left yet
14:43:45sjhillandersee: why is 'bits/ipc.h' only defined for alpha,mips,powerpc,sparc?
14:45:26sjhillit could be made common, as the files are pretty much identical
14:46:05sjhilli've got 3 undefined members from IPC related structures
14:46:14sjhilloh well, now i'm leaving
17:53:31sjhillmjn3: around?
17:55:12mjn3sjhill: yeah... fighting trying to crosscompile X for the Zaurus... and not having much luck.  :-(
17:55:19sjhillewww
17:55:32sjhillmjn3: have you looked at 'ipc.h' and friends much?
17:56:57mjn3a little.  it was one of the things that was failing in the perl tests... erik rewrote that stuff to use the kernel headers right before the 0.9.15 release
17:57:21sjhillhmmm, stuff isn't quite right i'm afraid
17:57:47mjn3not entirely unsurprising i suppose... how is it failing?
17:59:08sjhillin 'struct ipc_perm' in my kernel headers, i have:
17:59:14sjhillstruct ipc_perm
17:59:14sjhill  {
17:59:14sjhill    __key_t __key;                      /* Key.  */
17:59:14sjhill    __uid_t uid;                        /* Owner's user ID.  */
17:59:14sjhill    __gid_t gid;                        /* Owner's group ID.  */
17:59:14sjhill    __uid_t cuid;                       /* Creator's user ID.  */
17:59:17sjhill    __gid_t cgid;                       /* Creator's group ID.  */
17:59:20sjhill    unsigned short int mode;            /* Read/write permission.  */
17:59:22sjhill    unsigned short int __pad1;
17:59:24sjhill    unsigned short int __seq;           /* Sequence number.  */
17:59:26sjhill    unsigned short int __pad2;
17:59:28sjhill    unsigned long int __unused1;
17:59:30sjhill    unsigned long int __unused2;
17:59:32sjhill  };
18:01:21sjhillin the uClibc 'libc/sysdeps/linux/common/bits/ipc.h' we have:
18:01:28sjhill{
18:01:28sjhill    __kernel_key_t      key;
18:01:28sjhill    __kernel_uid_t      uid;
18:01:29sjhill    __kernel_gid_t      gid;
18:01:29sjhill    __kernel_uid_t      cuid;
18:01:29sjhill    __kernel_gid_t      cgid;
18:01:31sjhill    __kernel_mode_t     mode;
18:01:33sjhill    unsigned short      seq;
18:01:35sjhill};
18:01:45sjhillLTP won't compile because it wants the first structure definition
18:02:04sjhillbut it appears, the uCblic 'ipc.h' is getting priority
18:02:27sjhillperhaps 'ipc.h' should be removed from uClibc i think
18:03:40mjn3when you say in your kernel headers, do you mean the mips headers?
18:04:17sjhilli testing under x86 currently (because it's quicker)...but yes, the x86 kernel headers pointed to by the 'Config' file
18:06:49mjn3well, when erik changed things, he supposedly changed them to match the kernel struct defs for each arch.  on the other hand, glibc uses the same struct def on all archs and translates by copying individual struct members...  what is LTP doing that depends on the struct layout?
18:07:47sjhill#ifdef _LINUX_
18:07:47sjhill                printf ("seq    %d\n",  buf.msg_perm.__seq);
18:07:47sjhill                printf ("key    0x%x\n",  buf.msg_perm.__key);
18:07:47sjhill#else
18:07:48sjhill                printf ("seq    %d\n",  buf.msg_perm.seq);
18:07:48sjhill                printf ("key    %\n",  buf.msg_perm.key);
18:07:52sjhill#endif
18:07:54sjhill                
18:08:10sjhillthe __seq,__key members don't exist
18:10:14mjn3ahh... ok.  give me a couple of minutes to look at things.
18:10:34sjhilli can hack it to work, but probably not nicely
18:12:00sjhillmjn3: actually, it hink just 'libc/sysdeps/linux/common/bits/ipc.h' needs to be changed
18:41:01mjn3sjhill: sorry... just got off the phone.  you're probably right.  i just wanted to check and see if SUSv3 said anything about the members of struct ipc_perm
18:42:32sjhillk, i've already fixed the two header files so that things work
18:42:40sjhilli haven't checked them in yet
18:44:29mjn3seq and key aren't mentioned by SUSv3 as members of struct ipc_perm so i don't see any reason not to commit your changes.  key should probably be prefixed as well in common/bits/ipc.h
18:45:26mjn3is babbling... forget that last sentence.  ;-)
18:45:30sjhilli only changed files in 'common/bits'...the architecture specific files are correct, but since i'm doing x86 and there were not ipc.h files
18:45:41mjn3right
18:45:42sjhilli just changed the common bits
18:45:56sjhillcool, i'll check them in shortly...thx for your comments
18:46:03mjn3sure.
19:38:44anderseesjhill: memalign is back
19:39:01anderseesjhill: care to test and be sure I got it right?
19:39:26sjhillandersee: sure
19:42:40sjhillis very pleased how things have been progressing lately
19:42:56andersee:-)
19:46:34sjhillLTP flawlessly compiles against uClibc....valloc w/malloc-930716 works also...excellent!
19:46:52sjhillall of my patches to LTP were accepted also and will be in the next LTP release
19:46:58sjhillyippee!
19:47:29sjhillall that was for x86, now to do mips
19:47:43sjhillthx andersee
19:51:59sjhill*sigh* looks like i have a fair amount of work to get mips working
19:55:44anderseesjhill: cool.  :)
19:56:11anderseesjhill: about it all working (not about there being lots of work for mips)
19:58:14sjhillandersee: cool, the only failure was undefined 'syscall' for mips
20:01:59anderseesjhill: very cool.
20:02:11anderseesjhill: And syscall() can be fairly easily implemented.
20:16:17sjhillandersee: does 'syscall' always take 5 args?
20:16:37anderseesyscall takes as many args as are appropriate
20:16:52anderseeCould be 7 or 8 on some arches
20:16:56sjhillsmacks head on desk
20:17:03anderseeCould be no args depending on the syscall
20:17:04sjhillok, time for some research
20:18:05sjhillso this '__uClibc_syscall.S' is just a placeholder you made, correct?
20:18:13sjhilli remember you telling me that MIPS was a bit confusing
20:20:50sjhilloh wonderful, glibc has an elegant implementation of syscall for mips
20:20:54sjhilltries it
20:24:41anderseesjhill: the __uClibc_syscall.S is designed for a special unified syscall thing that I've mostly killed by implementing syscalls better
20:29:55sjhillis uberhacker
20:30:28anderseehehe
20:30:32anderseeSo it works?
20:30:44sjhillit compiles, which is good enough for starts :)
20:32:40sjhillbooting mips board
20:43:41sjhillkick ass
20:43:50sjhill'syscall' works on mips
20:50:27sjhillandersee: everything is checked in and working
20:50:34sjhillwhat an awesome day
20:52:27andersee:-)
20:53:08anderseeI give your boss permission to give you the rest of the day off.  :)
20:53:10sjhilltomorrow, 'lmbench' is next and then torture testing
20:53:18sjhillheh
21:45:40anderseeis away: I'm busy

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with infobot logs, split per channel and by date, etc.