#linuxcnc-devel | Logs for 2012-04-06

Back
[00:02:40] -!- skunkworks__ has quit [Ping timeout: 252 seconds]
[00:05:52] -!- rob_h has quit [Ping timeout: 245 seconds]
[00:06:29] -!- tom3p has quit [Quit: Ex-Chat]
[00:11:00] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[00:28:30] -!- kb8wmc [kb8wmc!~chatzilla@nat.mtp.cmsinter.net] has joined #linuxcnc-devel
[00:30:23] -!- Tom_L has quit []
[00:45:07] -!- Tom_itx has quit [Disconnected by services]
[00:45:13] Tom_garage is now known as Tom_itx
[01:19:44] -!- Nick001 has quit [Ping timeout: 246 seconds]
[01:19:57] -!- maximilian_h [maximilian_h!~bonsai@130.255.104.157] has joined #linuxcnc-devel
[01:22:33] -!- micges has quit [Quit: Leaving]
[01:26:45] -!- asdfasd has quit [Ping timeout: 248 seconds]
[02:00:01] -!- Tom_L has quit []
[02:02:01] -!- Tom_itx has quit [Disconnected by services]
[02:02:05] Tom_garage is now known as Tom_itx
[02:02:26] -!- zlog has quit [Read error: Operation timed out]
[02:16:40] -!- WCMD has quit [Quit: Page closed]
[02:19:59] -!- WCMD has quit [Client Quit]
[02:32:21] -!- Nick001 has quit [Ping timeout: 252 seconds]
[02:49:56] -!- Tom_L has quit []
[02:50:40] -!- jthornton has quit [Read error: Connection reset by peer]
[02:50:40] -!- JT-Shop has quit [Read error: Connection reset by peer]
[02:50:57] -!- koax [koax!~kelb@p508D5C17.dip.t-dialin.net] has joined #linuxcnc-devel
[02:51:06] -!- JT-Shop [JT-Shop!~John@184.20.140.167] has joined #linuxcnc-devel
[02:51:06] -!- jthornton [jthornton!~john@184.20.140.167] has joined #linuxcnc-devel
[02:54:26] -!- koax_ has quit [Ping timeout: 265 seconds]
[03:16:09] -!- pfred1 has quit [Quit: I'm off like a prom dress!]
[03:33:56] <mozmck1> GoSebGo: what did you do with the notes I sent you? did they get posted somewhere?
[03:40:18] -!- ve7it has quit [Remote host closed the connection]
[03:43:09] <seb_kuzminsky> i have them here but i haven't posted them anywhere
[03:43:35] <seb_kuzminsky> mozmck1: ^^^^
[03:44:04] <mozmck1> ok, me neither. I just noticed micges asked about it.
[03:47:32] <mozmck1> I think joe9 has built a pretty good kernel using 2.6.38.8, but it may be customized for his computer.
[03:47:50] <seb_kuzminsky> oh, did he get it to run? great
[03:48:55] <seb_kuzminsky> ok, here are the notes you sent me: http://highlab.com/~seb/linuxcnc/kernel/
[03:57:27] <mozmck1> Do you think they could go on the wiki? I might put them there later if I have some time.
[04:06:57] <seb_kuzminsky> mozmck1: i think they could, but i don't think that's the best place for them
[04:07:14] <seb_kuzminsky> be better to put them in a repo, along with the sources and scripts and stuff
[04:09:24] <mozmck1> ok
[04:45:48] -!- phantoxe has quit []
[04:52:51] -!- Radium has quit [Ping timeout: 265 seconds]
[04:54:40] -!- psha[work] [psha[work]!~psha@195.135.238.205] has joined #linuxcnc-devel
[05:03:16] -!- Valen has quit [Quit: Leaving.]
[05:07:20] -!- Nick001 has quit []
[05:41:25] e-ndy|afk is now known as e-ndy
[05:43:09] -!- Justin_CNC has quit [Quit: Leaving]
[06:05:52] -!- morfic has quit [Remote host closed the connection]
[06:07:45] -!- mhaberler [mhaberler!~mhaberler@195.191.253.94] has joined #linuxcnc-devel
[06:56:27] e-ndy is now known as e-ndy|afk
[06:59:41] e-ndy|afk is now known as e-ndy
[07:12:56] e-ndy is now known as e-ndy|afk
[07:15:28] e-ndy|afk is now known as e-ndy
[07:18:21] -!- Thetawaves has quit [Quit: This computer has gone to sleep]
[07:30:10] -!- capricorn_one has quit [Remote host closed the connection]
[07:49:27] -!- gene_ has quit [Remote host closed the connection]
[08:00:57] -!- kb8wmc has quit [Quit: ChatZilla 0.9.88.1 [Firefox 11.0/20120310193829]]
[08:14:31] -!- micges [micges!~micges@djm86.neoplus.adsl.tpnet.pl] has joined #linuxcnc-devel
[08:45:06] -!- rob__H [rob__H!~rob_h@5ace70c9.bb.sky.com] has joined #linuxcnc-devel
[09:05:25] -!- mhaberler has quit [Quit: mhaberler]
[09:05:37] -!- robh__ [robh__!rob_h@31.6.27.84] has joined #linuxcnc-devel
[09:06:05] -!- Jymmm has quit [Ping timeout: 252 seconds]
[09:09:05] -!- rob__H has quit [Ping timeout: 252 seconds]
[09:12:13] -!- robh__ has quit [Ping timeout: 276 seconds]
[10:05:59] -!- demacus_ has quit [Ping timeout: 245 seconds]
[10:58:33] AndroUser is now known as ebe23
[11:01:20] -!- ebe23 has quit [Client Quit]
[11:03:08] -!- mhaberler [mhaberler!~mhaberler@195.191.253.94] has joined #linuxcnc-devel
[11:04:19] -!- syyl has quit [Ping timeout: 272 seconds]
[12:01:46] -!- ktchk has quit [Read error: Connection reset by peer]
[12:13:11] -!- ebe23 has quit [Ping timeout: 246 seconds]
[12:20:44] -!- micges has quit [Quit: Leaving]
[12:30:22] <joe9> mozmck1: what do you need adeos for? i do not remember installing it, though.
[12:32:26] <jepler> a part of the rtai patches for the linux kernel are "adeos".
[12:32:33] <joe9> mozmck1: seb_kuzminsky: these are my notes, not sure if they can be of use to someone.
[12:32:54] <joe9> mozmck1: seb_kuzminsky: http://codepad.org/omNeCkoF
[12:33:02] <joe9> jepler: ok, thanks.
[12:33:35] <joe9> mozmck1: seb_kuzminsky: this is with 2.6.38.8 + vulcano
[12:39:59] -!- skunkworks [skunkworks!~skunkwork@68-115-41-210.static.eucl.wi.charter.com] has joined #linuxcnc-devel
[12:40:10] <skunkworks> logger[mah]:
[12:40:10] <logger[mah]> skunkworks: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc-devel/2012-04-06.html
[12:42:29] -!- rob_h [rob_h!rob_h@2002:1f06:16a8::1f06:16a8] has joined #linuxcnc-devel
[12:51:50] -!- ebe23 has quit [Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )]
[12:55:18] -!- mhaberler has quit [Quit: mhaberler]
[13:06:30] -!- syyl_ws has quit [Ping timeout: 244 seconds]
[13:07:00] -!- syyl_ has quit [Ping timeout: 252 seconds]
[13:32:03] -!- mk0 has quit [Quit: LinuxCNC 2.5.0!]
[13:32:40] -!- mhaberler [mhaberler!~mhaberler@195.191.253.94] has joined #linuxcnc-devel
[13:41:00] -!- rob_h has quit [Read error: Connection reset by peer]
[13:41:18] -!- rob_h [rob_h!rob_h@31.6.22.168] has joined #linuxcnc-devel
[13:50:27] -!- psha[work] has quit [Quit: Lost terminal]
[13:54:42] <seb_kuzminsky> joe9: looks like you did some good detective work there
[13:55:54] <seb_kuzminsky> i'll add your notes to that directory with mozmck's
[13:59:26] <cradek> I wonder how much of the kernel patch we really need. I think we only use one rtai feature...
[14:01:21] <mozmck1> I don't know, but the changes between the patch for 2.6.38.8 and the precise kernel look fairly non-trivial. I started through the code and figured out some of it, but others will take a lot more studying for me.
[14:01:35] -!- joe9 has quit [Quit: leaving]
[14:02:11] <mozmck1> someone who knows the kernel code better (wouldn't take much) could probably figure it out a lot quicker.
[14:02:31] <cradek> I tried the same (briefly)
[14:05:04] <jepler> it's enough to make one want to abandon rtai and all the software stepgen users, like some have advocated.
[14:05:49] <cradek> not sure if serious...
[14:05:56] <seb_kuzminsky> i hope not
[14:06:20] <seb_kuzminsky> are there alternatives to rtai that can give us reliable 1000 us scheduling for the servo thread? is preempt-rt there yet?
[14:06:21] <cradek> I really suspect that's 90% of our users.
[14:06:51] <mozmck1> Has anyone done much testing with the RT-preempt kernel? Looks like it has some support?
[14:07:10] <mozmck1> oh, seb just asked that.
[14:07:41] <seb_kuzminsky> well, off to do the kind of work that puts food on the table
[14:07:45] <seb_kuzminsky> see y'all later
[14:08:02] <mozmck1> aye, same here, except I'm doing it at home right now...
[14:08:32] -!- Loetmichel has quit [Ping timeout: 252 seconds]
[14:08:40] <cradek> is kernel.org 3.2.6 the most like the precise kernel?
[14:09:35] <mozmck1> no, it's up to 3.2.12 I think...
[14:09:54] <cradek> thanks
[14:10:26] <seb_kuzminsky> cradek: you could always git clone git://kernel.ubuntu.com/ubuntu/ubuntu-precise.git
[14:10:30] <seb_kuzminsky> ok NOW i'm leaving
[14:12:02] <mozmck1> I have it cloned, let me check. It was at 3.2.11 so I was actually using the vanilla kernel of that version for my work.
[14:13:45] <mozmck1> 3.2.12 as of a few days ago anyhow.
[14:15:18] <cradek> thanks
[14:17:43] <jepler> http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-precise.git;a=summary has 'rebase to v2.3.14' as a recent commit
[14:18:48] <jepler> I must not understand their workflow, though, because all that's in that commit is a change to debian/changelog
[14:20:48] -!- psha [psha!~psha@188.65.232.21] has joined #linuxcnc-devel
[14:21:43] <GoSebGo> jepler: the rebase itself wont show up as a commit in the history, right? So that's just a marker saying what they did
[14:22:12] <GoSebGo> Their rebased changes should ommediately preceed that commit
[14:22:15] <jepler> if that's really what they're doing it sure makes their "published" history useless
[14:23:01] <GoSebGo> Yes... Rebas destroys continuity-of-history. :-/
[14:23:06] <jepler> "rebase to v2.3.13" is a little further back in history, but now if history's really linear (and I can't see that from just gitweb) the tree at that commit is nonsense
[14:26:06] <mozmck1> yeah, I just did another pull and the makefile now says 3.2.14
[14:29:33] <mozmck1> anyone know why when I run 'git pull' on the precise kernel it often gives all kinds of merges and conflicts?
[14:30:01] <mozmck1> could it be because they are using a newer version of git?
[14:30:39] <mozmck1> (I haven't changed or committed anything)
[14:31:39] <cradek> it sounds like it's because they're idiots and use git in the stupidest way possible
[14:31:42] -!- Dimitrios has quit [Client Quit]
[14:33:11] <mozmck1> That inspires confidence in their kernel!
[14:33:35] <jepler> mozmck1: that would be another result of rebasing
[14:33:36] <mozmck1> Comes from the same company that created Unity after all I guess....
[14:34:22] <jepler> see the big section 'RECOVERING FROM UPSTREAM REBASE' in man git-rebase
[14:36:31] <mozmck1> interesting
[14:40:00] <cradek> https://wiki.ubuntu.com/Kernel/Dev/KernelGitGuide#Maintaining_local_changes
[14:41:59] <cradek> heh, second hit for ubuntu kernel git rebase wtf
[14:46:53] <GoSebGo> I wonder why they chose to do it that way
[14:47:49] <GoSebGo> I bet it was an agonized decision, i think they're neither stupid nor malicious
[14:48:16] <cradek> yeah, I suppose you're right.
[14:48:41] <jepler> Don't ascribe to stupidity or malice what can be ascribed to the actions of intelligent, good-intentioned people?
[14:48:47] <jepler> (well-intentioned?)
[14:49:26] <cradek> on the other hand, jepler has a point too.
[14:49:52] <cradek> who knows. looks like an awful decision from here.
[14:50:27] <jepler> if they never rebase then they end up carrying a change that is eventually accepted upstream on "their" side too
[14:51:16] <GoSebGo> Agreed, it's annoying for folks who want to keep local changes
[14:53:15] <GoSebGo> It forces everyone downstream (like us, if we ever get our shit together) to use the rebase workflow too
[14:53:55] <cradek> or just not use their git, which is how I feel it would be best to deal with them
[15:03:04] <GoSebGo> Use the linus git instead?
[15:04:59] <cradek> maybe. or since our shit, if we can get it together, would be one patch, just don't bother.
[15:05:47] <cradek> worrying about git is a distraction from worrying about rtai
[15:07:46] -!- Valen has quit [Quit: Leaving.]
[15:10:25] Cylly is now known as Loetmichel
[15:16:52] -!- syyl_ws has quit [Remote host closed the connection]
[15:20:42] -!- steves_logging has quit [Read error: Connection reset by peer]
[15:24:55] -!- steves_logging [steves_logging!~Steve@wsip-70-168-134-18.dc.dc.cox.net] has joined #linuxcnc-devel
[15:43:07] -!- mhaberler has quit [Quit: mhaberler]
[15:45:59] -!- linuxcnc-build has quit [Ping timeout: 246 seconds]
[15:46:04] -!- seb_kuzminsky has quit [Ping timeout: 265 seconds]
[15:46:11] -!- hm2-buildmaster has quit [Ping timeout: 252 seconds]
[15:47:45] -!- GoSebGo has quit [Quit: Bye]
[15:48:02] -!- GoSebGo [GoSebGo!~GoSebGo@184-229-35-132.pools.spcsdns.net] has joined #linuxcnc-devel
[15:48:49] <GoSebGo> Uh oh, not a good sign that the buildbot disconnected
[15:49:08] <GoSebGo> I wonder what just happened at my house
[15:58:58] -!- joe9 [joe9!~joe9@c-24-99-80-97.hsd1.ga.comcast.net] has joined #linuxcnc-devel
[16:07:45] e-ndy is now known as e-ndy|afk
[16:12:05] -!- linuxcnc-build [linuxcnc-build!~linuxcnc-@75-166-164-66.hlrn.qwest.net] has joined #linuxcnc-devel
[16:12:06] -!- seb_kuzminsky [seb_kuzminsky!~seb@75-166-164-66.hlrn.qwest.net] has joined #linuxcnc-devel
[16:12:23] -!- hm2-buildmaster [hm2-buildmaster!~hm2-build@75-166-164-66.hlrn.qwest.net] has joined #linuxcnc-devel
[16:13:36] <GoSebGo> Wow, linux preempt-rt claims latencies in the low tens of microseconds, about like rtai
[16:15:18] -!- joe9 has quit [Ping timeout: 272 seconds]
[16:17:47] -!- joe9 [joe9!~joe9@c-24-99-80-97.hsd1.ga.comcast.net] has joined #linuxcnc-devel
[16:19:52] -!- pjm__ has quit [Quit: TTFO]
[16:24:04] <GoSebGo> Wow, i'm getting all hot and bothered about preempt-rt now
[16:25:36] <GoSebGo> There's even a patch for 3.2.13
[16:26:11] <mozmck1> http://www.bitmuster.org/projects/emc.html
[16:27:02] <mozmck1> maybe it won't be too hard to get linuxcnc to use preempt-rt?
[16:29:59] -!- Radium has quit []
[16:32:19] <GoSebGo> Looks like jepler got it to run already, a while ago
[16:33:26] -!- GoSebGo has quit [Quit: Bye]
[16:33:41] -!- GoSebGo [GoSebGo!~GoSebGo@184-229-35-132.pools.spcsdns.net] has joined #linuxcnc-devel
[16:35:24] -!- Hydrogen has quit [Client Quit]
[16:35:42] <psha> GoSebGo: there is even linux-image-rt in stock debian :)
[16:36:23] <GoSebGo> Neat :-)
[16:47:37] <psha> probably i'll try it for our trading system :)
[16:49:01] -!- Hydrogen has quit [Client Quit]
[16:51:00] -!- psha has quit [Quit: Lost terminal]
[16:58:15] -!- ve7it [ve7it!~LawrenceG@S0106001c10b7770f.pk.shawcable.net] has joined #linuxcnc-devel
[16:58:43] <skunkworks> I thought it was probably good enough for servo machines - not fast enough for software step gen
[17:02:32] <GoSebGo> skunkworks: i thought so too, but here is a latency result table that shows software-stepgen-capable latencies: https://rt.wiki.kernel.org/articles/c/o/n/CONFIG_PREEMPT_RT_Patch_79df.html#Platforms_Tested_and_in_Use_with_CONFIG_PREEMPT_RT
[17:10:47] <skunkworks> wow - those numbers do look pretty good
[17:13:28] <skunkworks> https://www.osadl.org/Continuous-latency-monitoring.qa-farm-monitoring.0.html
[17:15:39] <skunkworks> looks like Intel Intel i7-2600K @3400 MHz, 64 bit is the best there... :)
[17:17:35] -!- kb8wmc [kb8wmc!~chatzilla@nat.mtp.cmsinter.net] has joined #linuxcnc-devel
[17:17:45] -!- Radium_ has quit [Client Quit]
[17:23:28] <awallin> i7-2600k has been the "most-bang-for-the-buck" desktop processor for a long time now IMO
[17:24:09] <GoSebGo> A good cpu is good, but not enough for low latency
[17:24:46] <GoSebGo> You need the right supporting chipset and firmware on the motherboard, and the right peripherals/drivers
[17:30:46] -!- GoSebGo has quit [Quit: Bye]
[17:31:03] -!- GoSebGo [GoSebGo!~GoSebGo@184-229-35-132.pools.spcsdns.net] has joined #linuxcnc-devel
[17:33:50] <awallin> skunkworks: how do those latency-plots you linked to relate to the "max jitter" number we are more used to ?
[17:41:32] e-ndy|afk is now known as e-ndy
[17:41:47] -!- micges [micges!~micges@djm86.neoplus.adsl.tpnet.pl] has joined #linuxcnc-devel
[17:47:02] <micges> mozmck1: hi
[17:48:05] <micges> I've been following your rtai building notes
[17:48:45] <micges> everything is ok except that rtai-modules won't build
[17:49:38] -!- mhaberler [mhaberler!~mhaberler@195.191.253.94] has joined #linuxcnc-devel
[17:53:19] -!- psha [psha!~psha@213.208.162.69] has joined #linuxcnc-devel
[17:56:59] <micges> (ignore above, saying at loud often helps)
[18:04:46] <skunkworks> heh
[18:04:51] <skunkworks> awallin: no clue..
[18:09:29] -!- Tom_itx has quit [Disconnected by services]
[18:09:32] Tom_garage is now known as Tom_itx
[18:10:08] -!- zlog has quit [Ping timeout: 240 seconds]
[18:20:09] <mozmck1> micges: I know the feeling (saying things out loud...)
[18:20:48] <micges> :)
[18:21:53] -!- Tom_L has quit []
[18:26:04] <awallin> skunkworks: ok, It looks like the minimum latency they get in those plots is around 10us, and when there is a delay in the system they sometimes get 50us. That would be a jitter of 50-10=40us. I think to compare to the linuxcnc latency-test one would want a histogram centered around e.g. 1ms (for servo thread) or faster for base-thread.
[18:26:53] -!- skunkworks has quit []
[18:29:31] -!- dgarr [dgarr!~dgarrett@adsl-75-61-68-165.dsl.pltn13.sbcglobal.net] has joined #linuxcnc-devel
[18:29:58] <dgarr> module_helper/module_helper.c:58: error: ‘return’ with a value, in function returning void
[18:30:10] -!- micges has quit [Ping timeout: 265 seconds]
[18:31:44] <cradek> dgarr: what branch?
[18:32:55] <dgarr> master: d247f8f8
[18:35:30] <cradek> I suspect he meant exit(1)
[18:35:48] <cradek> strongly suspect
[18:38:05] <awallin> is module_helper.c only built for realtime? I don't see that error with --enable-simulator...
[18:38:17] <cradek> yes
[18:38:25] <cradek> it's the setuid program that inserts/removes kernel modules
[18:40:46] -!- IchGuckLive has quit [Quit: ChatZilla 0.9.87 [Firefox 10.0.2/20120216080748]]
[18:42:42] -!- micges [micges!~micges@efk85.neoplus.adsl.tpnet.pl] has joined #linuxcnc-devel
[18:45:40] -!- ben_says has quit [Quit: Leaving]
[18:52:17] -!- ben_says has quit [Quit: Leaving]
[18:54:53] -!- ben_says has quit [Remote host closed the connection]
[18:57:14] -!- Radium has quit [Ping timeout: 245 seconds]
[19:02:02] -!- micges has quit [Quit: Leaving]
[19:02:04] <jepler> argh
[19:02:16] <jepler> sorry all
[19:03:19] <CIA-68> 03jepler 07master * r291fc50aa627 10/src/module_helper/module_helper.c: module_helper: fix compile error I introduced
[19:04:05] <jepler> I'll carry a rubber weevil as the mark of my shame
[19:05:42] <jepler> Tracker had 9 seeders and 1 leechers 7 minutes ago
[19:05:56] <jepler> well the linuxcnc iso tracker is not well-populated, but at least it is working
[19:06:03] <jepler> I've personally uploaded less than 1 iso worth so far
[19:16:17] -!- micges [micges!~micges@efk85.neoplus.adsl.tpnet.pl] has joined #linuxcnc-devel
[19:19:03] <dgarr> i've done some testing with rt_preempt: http://www.panix.com/~dgarrett/stuff/rt_preempt/notes.rt.txt
[19:31:39] -!- psha has quit [Quit: Lost terminal]
[19:32:23] -!- mrsun has quit [Quit: Leaving]
[19:40:51] -!- vladimirek has quit [Ping timeout: 252 seconds]
[19:41:18] <jepler> do any hardware drivers work with those patches and --enable-realtime=linux?
[19:45:06] <jepler> at least some code duplication in linux_rtapi_app.cc should be avoidable
[19:46:51] <micges> jepler: in short way, those setup will run linuxcnc simulator with very predictible latency?
[19:47:57] <micges> becouse our drivers don't compile with other than RTAI
[19:49:06] <jepler> our drivers currently make lots of assumptions that the are running in kernel space (call linux kernel APIs freely during setup and shutdown)
[19:49:17] <jepler> s/that the/that they/
[19:50:01] <micges> yeah I was talking about that
[19:50:07] <jepler> so before rt-preempt-based-on-"sim" is useful something has to be done about this .. I was inquiring whether the patch dgarr mentioned began to do any of that
[19:50:47] <jepler> I didn't see it in the patch, but maybe I overlooked something..
[19:51:12] <micges> no, drivers are not touched
[19:53:02] -!- adb has quit [Ping timeout: 250 seconds]
[19:53:27] <micges> another thing, drivers should be preemptible, as whole kernel is, right?
[19:55:58] <micges> but I assume they are becouse two or more rt threads doesn't make any problems
[19:59:45] <jepler> I don't know what guarantees you can get about the priority of the rtapi tasks (which are implemented as linux userspace threads here) compared to other userspace programs or to kernel space
[20:02:54] <jepler> sched_setscheduler(SCHED_DEADLINE or SCHED_FIFO) are intended to set the scheduling priority of rtapi_app
[20:10:16] -!- adb [adb!~adb@178-211-231-163.dhcp.voenergies.net] has joined #linuxcnc-devel
[20:20:09] -!- demacus has quit [Ping timeout: 245 seconds]
[20:28:09] <jepler> also imo there'd be no reason to carry around two userspace rtapi implementations; if this "new" one---which is mostly a copy of sim---works without serious caveats for non-hardware-controlling setups, then I'd say to just drop "sim"
[20:29:28] <jepler> maybe configure --enable-realtime=sim would remain separate and build without setuid but otherwise why not share all that code?
[20:36:17] -!- mhaberler has quit [Quit: mhaberler]
[20:37:14] <cradek> my shot in the dark (untested): http://timeguy.com/cradek-files/emc/hal-linux-3.2.14-x86-32only-cjr1.patch
[20:37:47] <cradek> there is one spot where I'm a bit unsure, and I skipped some x86_64-specific asm code, so don't try that.
[20:37:53] <jepler> you're a brave man
[20:38:05] <cradek> seems like I'll have to install precise to try building it
[20:39:37] <cradek> it's hard to distinguish bravery from foolishness and/or hubris.
[20:40:17] <mozmck1> cradek: is that a modified adeos patch?
[20:41:21] <cradek> yes
[20:41:46] <mozmck1> neat.
[20:42:12] <cradek> I may try building it this weekend sometime
[20:42:14] <mozmck1> I have precise installed on a second hard drive, but I'm pretty busy this weekend.
[20:42:54] -!- maximilian_h has quit [Quit: Leaving.]
[20:43:24] <mozmck1> did you grep to see if there were any new local_irq_save/restore functions that might need to be changed to the _hw versions?
[20:43:59] <cradek> no - is the pattern that all of them get changed?
[20:44:36] <mozmck1> I don't really know, but it's something I had thought of checking.
[20:44:50] <cradek> I see that's sure not the pattern
[20:45:15] <mozmck1> ok.
[20:46:11] <cradek> I had similar thoughts about ipipe_mm_switch_[un]protect
[20:46:57] -!- maximilian_h [maximilian_h!~bonsai@130.255.104.157] has joined #linuxcnc-devel
[20:47:49] <cradek> in particular unshare(2) lost a reference to mm (mmap/munmap sharing?) and that makes me nervous
[20:47:56] <mozmck1> yeah, I was going to check for other stuff. I'm pretty sure ipipe should intercept all interrupts
[20:48:26] <mozmck1> hmm, I don't know anything about that.
[20:48:46] <cradek> feels like there's a 10% I got everything right. I don't have high hopes but am sharing in case anyone wants to compare notes.
[20:49:19] <mozmck1> that's on a vanilla kernel?
[20:49:22] <cradek> yes
[20:49:50] -!- DJ9DJ has quit [Quit: bye]
[20:49:59] <cradek> I figured it'd be wise to tackle ubuntu's changes and debian building separately
[20:50:21] <mozmck1> I'll try and look at it some. Company coming this weekend, real busy with work, and our first baby is just over 2 months old now and still not sleeping all night.
[20:50:40] <cradek> yikes!
[20:51:20] <cradek> sounds like the perfect time to mess with kernel patches.
[20:53:21] <mozmck1> :) maybe at 2 in the morn if I can stay awake!
[20:55:17] <mozmck1> How did you create the patch? Did you have a separate dir with a clean kernel?
[20:55:27] <cradek> yes
[20:55:50] <mozmck1> Ok, just curious because I haven't done much of that.
[20:55:56] e-ndy is now known as e-ndy|afk
[21:01:26] -!- FinboySlick has quit [Quit: Leaving.]
[21:04:05] -!- bedah has quit [Quit: bye]
[21:18:52] -!- micges has quit [Quit: Leaving]
[21:39:36] -!- jepler has quit [Quit: Reconnecting]
[21:40:29] -!- jepler [jepler!~jepler@emc/developer/pdpc.professional.jepler] has joined #linuxcnc-devel
[21:41:17] jepler is now known as jepler_
[21:41:23] jepler_ is now known as jepler
[21:53:10] -!- Fox_Muldr has quit [Ping timeout: 265 seconds]
[21:55:11] syyl__ is now known as syyl
[22:14:50] -!- Beryllium has quit []
[22:26:56] -!- kbarry has quit [Quit: ChatZilla 0.9.88.1 [Firefox 11.0/20120312181643]]
[22:27:18] -!- syyl has quit [Quit: Leaving]
[23:01:06] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[23:47:57] -!- Himanshu has quit [Quit: Page closed]