Back
[00:00:01] <andypugh> micges: Perhaps you need to disable the TRAM stuff?
[00:00:03] <micges> andypugh: relax, after it will work I'll enable sserial to optimize it slightly
[00:00:07] <seb_kuzminsky> 75 reads/invocation seems pretty reasonable
[00:00:21] <seb_kuzminsky> the tram stuff shouldnt affect the number of i/o operations
[00:00:38] <seb_kuzminsky> it only saves you a bunch of writes when using epp
[00:00:42] <seb_kuzminsky> pci isnt affected
[00:01:29] <andypugh> I was unclear.
[00:01:54] <andypugh> I was meaning not to start the automatic reads/writes until good and ready
[00:01:59] -!- Brandonian has quit [Quit: Brandonian]
[00:02:00] <seb_kuzminsky> oh
[00:03:03] <seb_kuzminsky> what's the sserial interface like? i haven't looked at it
[00:03:22] <seb_kuzminsky> there's a module in the fpga that's a buffer to a uart?
[00:03:41] <andypugh> 2 registers per card, N-registers per module. All get read and written every cycle
[00:04:13] <andypugh> (I mean N registers per sserial remote)
[00:04:45] <andypugh> Yes, rather than me describing it, it is perhaps easier to take a look at regmap
[00:04:52] <seb_kuzminsky> heh ok
[00:05:24] <andypugh> Very nearly all the heavy-lifting is done by the FPGA.
[00:05:40] <seb_kuzminsky> that sounds good :-)
[00:06:17] <andypugh> At load-time the remotes say what all the bits in their 3 32-bit registers mean.
[00:06:38] -!- logger[mah] has quit [Remote host closed the connection]
[00:06:44] -!- logger[mah] [logger[mah]!~loggermah@mail.mah.priv.at] has joined #linuxcnc-devel
[00:07:10] <andypugh> (a nice way to see the structure is in sserial.h which has psuedo-structures for the 8i20 and 7i64, because they predated the clever stuff)
[00:08:10] -!- mhaberler [mhaberler!~mhaberler@extern-181.stiwoll.mah.priv.at] has joined #linuxcnc-devel
[00:08:32] <andypugh> {but only in master [I am running out of levels of parenteses]}
[00:08:53] <seb_kuzminsky> you look like a configure script
[00:09:38] <micges> with 7i80 and rtnet there is another big problem - rtnet is designed to be used in userspace rt task
[00:10:08] <micges> like in xenomai
[00:10:17] <andypugh> That might be OK if we move to Xenomai.
[00:11:00] <andypugh> You could just assume that the 7i80 and Xenomai support will arrive in the same release?
[00:12:07] -!- ybon has quit [Ping timeout: 245 seconds]
[00:12:35] <micges> I want it also under RTAI
[00:13:04] <andypugh> Why?
[00:14:45] <andypugh> (That wasn't meant to sound rude). My (limited) understanding is that Xenomai Userspace is currently showing unexpectedly low latency
[00:15:43] <micges> becaouse of stabilisation of lcnc under RTAI
[00:16:30] <andypugh> I am not saying it is a bad ambition, just that all is not necessarily lost if it proves too hard.
[00:17:51] <andypugh> Would it be possible to write your own RTAI ethernet driver, only using the device-specific parts of RTNet? (I have briefly looked at the possibility of doing that with COMEDI)
[00:18:53] <micges> nope
[00:19:15] <seb_kuzminsky> i, too, value all the runtime we have with rtai
[00:19:18] <seb_kuzminsky> bbl
[00:24:08] <andypugh> I confess I am quite comfortable with RTAI now.
[00:24:39] <micges> yes it will take a while to get comfortable with xenomai
[00:29:00] -!- factor has quit [Read error: Connection reset by peer]
[00:33:37] <andypugh> Do any other projects use RTAI?
[00:36:05] <PCW> Speed wise for initialization the driver should pack as may reads as possible per packet (and possibly read the remote database with 8 byte read commands)
[00:38:48] <PCW> another thing that would help would be a local polling engine for things like sserial because of the expense of polling remotely over Ethernet
[00:39:57] <PCW> I still dont see how you get 1.5 million reads, 10 thousand maybe
[00:41:12] <micges> me either
[00:43:32] <PCW> Maybe a bug, even in the worst case there are only maybe 1000 bytes of database data per sserial remote it may take 6 or so packets to get that byte but that only 6000 ops
[00:46:27] <andypugh> 1.5 million reads, even at 2.5Mb would take longer than I have ever noticed it taking too.
[00:47:16] <PCW> I just dont see how you get that many read or write ops
[00:48:53] <andypugh> Actually... I can imagine the reads being true..
[00:49:17] <PCW> (of course with PCI the happen in less than a usec put on Ethernet they will take maybe 50 usec (but maybe 1 ms if the driver is only invoked in the servo thread)
[00:49:18] <andypugh> I don't know if there is a sleep in sserial_wait_for
[00:50:11] <micges> maybe there is repeat somewhere if read failed?
[00:50:21] <PCW> yeah the polling is a pain (but the second Ethernet packet should always succeed)
[00:50:46] <andypugh> micges: Try putting an rtapi_wait in sserial.c somewhere around 1518
[00:50:52] -!- asdfas has quit [Ping timeout: 245 seconds]
[00:51:15] <micges> I'll try
[00:51:19] <PCW> so you poll slower but should have to wait no longer (+- polling granularity)
[00:51:20] <andypugh> That is, on reflection, rather too busy.
[00:53:05] <andypugh> <embarassed>
[00:53:51] <andypugh> A 2mS wait in there would probably cause no problems at all.
[00:54:22] <andypugh> It is only called by init code.
[00:56:17] <PCW> A remote read on sserisl should only take say 30 usec at most, Ethernet polling will be slower so should just reduce the number of poll calls
[00:56:29] <andypugh> That's a really short loop.
[00:57:22] <PCW> (But it still has to wait for the Ethernet xmit/recv turnaround)
[00:58:30] <PCW> hence the slower polling but the overall wait for sserial command complete should not be much longer
[00:59:32] <andypugh> There is no argument from me but that that section of code displays deadful noobism.
[01:00:22] -!- Nick001-Shop has quit [Remote host closed the connection]
[01:01:25] <PCW> micges: is it that the startup code only polls the Ethernet at 1KHz ?
[01:01:53] <micges> seems so
[01:02:13] -!- pjm has quit [Read error: Connection reset by peer]
[01:02:36] <andypugh> Unfortunately that bit of startup code (of mine) polls the cmd reg as fast as it possibly can. For no good reason.
[01:02:37] <PCW> you really want the Ethernet driver to be completely polled
[01:05:15] <andypugh> micges: Have you tried a wait in there yet?
[01:06:22] <PCW> Stll don't see how 1.5 million ops is possible
[01:06:29] <micges> not now, I'm in other part of code now
[01:07:40] <andypugh> PCW: I do.
[01:07:49] <PCW> How?
[01:08:12] <andypugh> It really is a very short loop checking the command reg for being zero as fast as possible.
[01:08:40] <andypugh> Direct access, nothing to do with threads.
[01:09:22] <andypugh> At that point the whole PC is doing nothing but checking the command reg...
[01:09:24] <micges> yes but each direct access is now eth read
[01:09:32] <PCW> but for the 7I80 that checking involves sending a data request packet over Ethernet and waiting for the response
[01:09:59] <PCW> the second request should always succeed
[01:10:07] <andypugh> Well, yes. It is probably worse with PCI devices. But nobody has noticed yet.
[01:11:10] <PCW> well in both cases you wait till ready so that should not impact the total time (and Ethernet should have fewer ops)
[01:11:53] <PCW> unless because of the way the Ethernet driver works its only polling at 1 KHz
[01:13:58] <PCW> a easy 8-one reduction in database read time would be to suck up the data in 64 bit chunks (instead of 8)
[01:14:11] <PCW> 8
[01:16:15] <PCW> a little uglier checking for the terminating nulls on the two strings at the end of the record
[01:16:57] <andypugh> Is it possble the over-enthusiastic polling fills a queue?
[01:17:41] <PCW> I was wondering about that with the Ethernet driver
[01:18:31] <PCW> Does sound like that kind of a problem
[01:20:00] <PCW> that is something that might explain 1.5 M packets
[01:22:11] <PCW> but that is something easy to check, the polling loop should only run a couple times
[01:23:32] <PCW> (with Ethernet because of the ~ 50 uSec turnaround) on PCI a remote read might poll for 100 times or so
[01:48:59] -!- ssi_ has quit [Ping timeout: 256 seconds]
[02:03:49] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[02:04:18] -!- sumpfralle has quit [Ping timeout: 245 seconds]
[02:05:17] -!- elmo40 has quit [Ping timeout: 252 seconds]
[02:05:21] -!- L33TG33KG34R has quit [Read error: Connection reset by peer]
[02:07:15] -!- alex_joni has quit [Ping timeout: 240 seconds]
[02:07:54] -!- alpha1125 has quit [Quit: Computer has gone to sleep.]
[02:08:29] -!- alex_joni [alex_joni!~alex_joni@emc/board-of-directors/alexjoni] has joined #linuxcnc-devel
[02:08:29] -!- mode/#linuxcnc-devel [+v alex_joni] by ChanServ
[02:13:29] -!- cevad has quit [Quit: Leaving]
[02:24:42] -!- andypugh has quit [Quit: andypugh]
[02:24:52] -!- micges has quit [Quit: Leaving]
[03:10:01] -!- FinboySlick has quit [Quit: Leaving.]
[03:20:14] zz_satyag is now known as satyag
[03:55:00] -!- mevon_ has quit [Quit: Ex-Chat]
[03:55:03] -!- factor has quit [Read error: Operation timed out]
[04:00:09] -!- Keknom has quit [Quit: Leaving.]
[04:06:08] -!- ve7it [ve7it!~LawrenceG@S0106001c10b7770f.pk.shawcable.net] has joined #linuxcnc-devel
[04:17:52] -!- ve7it has quit [Remote host closed the connection]
[04:21:07] -!- Loetmichel has quit [Ping timeout: 260 seconds]
[04:21:23] satyag is now known as zz_satyag
[04:24:02] -!- Aero-Tec has quit [Ping timeout: 255 seconds]
[04:24:10] Aero-Tec_ is now known as Aero-Tec
[04:36:11] -!- Newtonianb has quit [Ping timeout: 255 seconds]
[04:42:04] -!- mhaberler has quit [Quit: mhaberler]
[05:09:42] -!- kmiyashiro has quit [Quit: kmiyashiro]
[05:20:24] -!- karavanjoW has quit [Quit: KVIrc 4.0.4 Insomnia http://www.kvirc.net/]
[05:34:17] -!- vladimirek [vladimirek!~vladimire@95.105.250.72] has joined #linuxcnc-devel
[05:39:37] -!- RoyOnWheels has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
[05:56:10] -!- kwallace [kwallace!~kwallace@smb-236.sonnet.com] has parted #linuxcnc-devel
[05:57:41] -!- zzolo has quit [Quit: zzolo]
[06:01:55] -!- Fox_Muldr has quit [Ping timeout: 260 seconds]
[06:33:08] mazafaka1 is now known as mazafaka
[06:43:17] -!- psha [psha!~psha@213.208.162.92] has joined #linuxcnc-devel
[06:45:42] -!- Cylly has quit []
[07:13:54] -!- the_wench has quit [Ping timeout: 240 seconds]
[07:14:14] -!- archivist has quit [Ping timeout: 240 seconds]
[07:16:56] -!- dhoovie has quit [Ping timeout: 246 seconds]
[07:42:40] -!- bradsimantel has quit [Quit: bradsimantel]
[07:53:34] -!- herron has quit [Ping timeout: 240 seconds]
[08:29:50] -!- cncbasher has quit [Remote host closed the connection]
[08:30:20] -!- psha has quit [Quit: leaving]
[08:52:41] -!- tayy has quit [Remote host closed the connection]
[08:56:27] -!- Newtonianb has quit [Ping timeout: 244 seconds]
[09:14:24] theos is now known as Guest16295
[09:18:00] -!- Guest16295 has quit [Ping timeout: 276 seconds]
[09:26:07] -!- hm2-buildmaster [hm2-buildmaster!~hm2-build@174-16-205-244.hlrn.qwest.net] has joined #linuxcnc-devel
[09:26:23] -!- seb_kuzm1nsky [seb_kuzm1nsky!~seb@174-16-205-244.hlrn.qwest.net] has joined #linuxcnc-devel
[09:27:55] -!- linuxcnc-build_ has quit [Ping timeout: 255 seconds]
[09:28:11] -!- seb_kuzminsky has quit [Ping timeout: 260 seconds]
[09:28:14] -!- hm2-buildmaster_ has quit [Ping timeout: 255 seconds]
[09:29:07] -!- b_b has quit [Changing host]
[09:37:34] -!- hm2-buildmaster_ [hm2-buildmaster_!~hm2-build@174-16-209-159.hlrn.qwest.net] has joined #linuxcnc-devel
[09:37:54] -!- seb_kuzminsky [seb_kuzminsky!~seb@174-16-209-159.hlrn.qwest.net] has joined #linuxcnc-devel
[09:38:31] -!- hm2-buildmaster has quit [Ping timeout: 265 seconds]
[09:39:28] -!- seb_kuzm1nsky has quit [Ping timeout: 276 seconds]
[09:41:38] -!- linuxcnc-build [linuxcnc-build!~linuxcnc-@174-16-209-159.hlrn.qwest.net] has joined #linuxcnc-devel
[10:20:35] -!- cncbasher [cncbasher!~quassel@cpc15-hart9-2-0-cust101.11-3.cable.virginmedia.com] has joined #linuxcnc-devel
[10:25:07] -!- cncbasher has quit [Read error: Connection reset by peer]
[10:37:34] -!- rob_h [rob_h!~rob_h@027c00e2.bb.sky.com] has joined #linuxcnc-devel
[10:38:56] -!- tayy has quit [Remote host closed the connection]
[10:47:33] -!- The_Ball has quit [Read error: No route to host]
[10:52:26] zz_satyag is now known as satyag
[10:53:56] satyag is now known as zz_satyag
[10:54:52] zz_satyag is now known as satyag
[11:04:40] -!- dhoovie has quit [Read error: Connection reset by peer]
[11:19:58] -!- ybon has quit [Quit: WeeChat 0.3.8]
[11:28:15] satyag is now known as zz_satyag
[11:53:11] zz_satyag is now known as satyag
[12:15:09] -!- mhaberler [mhaberler!~mhaberler@extern-181.stiwoll.mah.priv.at] has joined #linuxcnc-devel
[12:27:47] -!- yuvipanda has quit [Client Quit]
[12:29:50] -!- factor has quit [Read error: Connection reset by peer]
[12:46:56] satyag is now known as zz_satyag
[12:48:35] -!- adb [adb!~IonMoldom@178-211-235-11.dhcp.voenergies.net] has joined #linuxcnc-devel
[12:52:34] -!- dway has quit [Client Quit]
[12:53:05] -!- horatio_cromwell has quit [Read error: Connection reset by peer]
[12:54:14] -!- asdfasd has quit [Ping timeout: 240 seconds]
[13:07:44] -!- plushy has quit [Quit: Leaving.]
[13:11:15] -!- holst has quit [Ping timeout: 260 seconds]
[13:19:11] -!- rob_h has quit [Read error: Connection reset by peer]
[13:19:35] -!- rob_h [rob_h!~rob_h@027c00e2.bb.sky.com] has joined #linuxcnc-devel
[13:30:26] -!- toastyde1th has quit [Read error: No buffer space available]
[13:34:30] -!- cmorley has quit [*.net *.split]
[13:34:31] -!- Theta9 has quit [*.net *.split]
[13:34:31] -!- mxn has quit [*.net *.split]
[13:34:31] -!- spiderdijon has quit [*.net *.split]
[13:34:31] -!- _ink has quit [*.net *.split]
[13:34:31] -!- Jymmm has quit [*.net *.split]
[13:34:32] -!- zod has quit [*.net *.split]
[13:43:03] zz_satyag is now known as satyag
[13:43:20] -!- cmorley [cmorley!~chris@S010600c09fc019c2.no.shawcable.net] has joined #linuxcnc-devel
[13:44:18] -!- ZinovaS has quit [Remote host closed the connection]
[13:44:34] -!- jst has quit [Read error: Connection reset by peer]
[13:45:11] -!- hewball has quit [Read error: Connection reset by peer]
[13:47:35] -!- icee has quit [Ping timeout: 240 seconds]
[13:53:48] zod_ is now known as zod
[14:06:25] -!- zzolo has quit [Quit: zzolo]
[14:19:52] -!- mxn_ has quit [Quit: leaving]
[14:20:06] -!- mxn has quit [Changing host]
[14:22:54] -!- mhaberler has quit [Quit: mhaberler]
[14:23:56] -!- zandzpider has quit [Read error: Connection reset by peer]
[14:27:43] -!- mhaberler [mhaberler!~mhaberler@macbook.stiwoll.mah.priv.at] has joined #linuxcnc-devel
[14:34:56] satyag is now known as zz_satyag
[14:37:36] -!- mackerski has quit [Ping timeout: 240 seconds]
[14:57:56] -!- holst has quit [Ping timeout: 248 seconds]
[14:59:10] -!- zzolo has quit [Quit: zzolo]
[14:59:43] -!- yoyoek1 has quit [Quit: Ex-Chat]
[15:15:12] -!- dway has quit [Quit: dway]
[15:21:06] -!- the_wench has quit [Ping timeout: 264 seconds]
[15:21:06] -!- archivist has quit [Ping timeout: 264 seconds]
[15:34:32] -!- abetusk has quit [Quit: Leaving]
[15:41:04] -!- pingufan has quit [Quit: Konversation terminated!]
[15:50:09] -!- dway has quit [Quit: dway]
[16:03:58] -!- dway has quit [Client Quit]
[16:06:19] -!- gentux has quit [Ping timeout: 255 seconds]
[16:11:47] -!- kmiyashiro has quit [Quit: kmiyashiro]
[16:19:37] -!- gentux_ has quit [Client Quit]
[16:22:07] -!- Aero-Tec has quit [Ping timeout: 245 seconds]
[16:22:12] Aero-Tec_ is now known as Aero-Tec
[16:22:36] -!- bmwyss has quit [Read error: Operation timed out]
[16:32:04] -!- bradsimantel has quit [Quit: bradsimantel]
[16:32:25] -!- plushy has quit [Ping timeout: 256 seconds]
[16:45:14] -!- alpha1125 has quit [Quit: Computer has gone to sleep.]
[16:50:26] -!- bmwyss has quit [Quit: bmwyss]
[16:59:55] -!- purejamie has quit [Quit: Page closed]
[17:04:21] -!- mackerski has quit [Ping timeout: 248 seconds]
[17:04:21] mackerski_ is now known as mackerski
[17:20:42] -!- Newtonianb has quit [Client Quit]
[17:20:43] -!- theorbtwo has quit [Read error: Connection reset by peer]
[17:24:35] -!- pjm has quit [Read error: Connection reset by peer]
[17:28:59] -!- psha [psha!~psha@213.208.162.92] has joined #linuxcnc-devel
[17:31:32] -!- wboykinm has quit [Remote host closed the connection]
[17:44:11] -!- cncbasher [cncbasher!~quassel@cpc15-hart9-2-0-cust101.11-3.cable.virginmedia.com] has joined #linuxcnc-devel
[17:55:54] -!- sumpfralle has quit [Ping timeout: 264 seconds]
[17:57:45] -!- ybon has quit [Quit: WeeChat 0.3.8]
[17:59:51] plushy1 is now known as plushy
[18:03:17] -!- pjm has quit [Read error: Connection reset by peer]
[18:03:44] -!- mackerski has quit [Quit: mackerski]
[18:31:50] -!- Jipeiro has quit [Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org]
[18:43:34] -!- tjb1 has quit [Ping timeout: 240 seconds]
[18:51:18] -!- jp_ has quit [Ping timeout: 276 seconds]
[18:52:27] -!- zz_satyag has quit [Remote host closed the connection]
[19:03:58] -!- ve7it [ve7it!~LawrenceG@S0106001c10b7770f.pk.shawcable.net] has joined #linuxcnc-devel
[19:20:30] -!- mhaberler has quit [Ping timeout: 264 seconds]
[19:35:10] -!- IchGuckLive has quit [Quit: ChatZilla 0.9.87 [Firefox 16.0.2/20121025205401]]
[19:52:23] -!- tjb1 has quit [Read error: Connection reset by peer]
[19:52:54] -!- ve7it has quit [Remote host closed the connection]
[19:54:39] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust639.basl.cable.virginmedia.com] has joined #linuxcnc-devel
[19:57:30] -!- alpha1125 has quit [Ping timeout: 252 seconds]
[19:57:34] -!- rob__H [rob__H!~rob_h@027c009b.bb.sky.com] has joined #linuxcnc-devel
[20:00:58] -!- rob_h has quit [Ping timeout: 256 seconds]
[20:04:35] -!- b_b has quit [Changing host]
[20:14:57] -!- psha has quit [Quit: Lost terminal]
[20:22:07] -!- bradsimantel has quit [Ping timeout: 245 seconds]
[20:22:07] bradsimantel_ is now known as bradsimantel
[20:23:54] <andypugh> What are the numbers in Dewey's post?
[20:39:58] <PCW> Andy: 7I73 encoder issue is a real bug (encmodeN not updated from nvencmodeN ), but you can
[20:40:00] <PCW> set the volatile emcmodeN
[20:44:07] <andypugh> Maybe I can try the firmware updater on the 7i73 first :-)
[20:44:37] <andypugh> (I might switch to resolvers for the spindle encoder anyway)
[20:48:43] <PCW> Glad you noticed, looks like we broke that and its one of the corner things we didnt test for
[20:51:19] -!- tjb1 has quit [Read error: Connection reset by peer]
[20:54:39] -!- zzolo has quit [Quit: zzolo]
[20:56:27] -!- bradsimantel has quit [Ping timeout: 252 seconds]
[20:59:57] -!- tjb1 has quit [Ping timeout: 244 seconds]
[20:59:57] tjb1_ is now known as tjb1
[21:07:35] -!- ve7it [ve7it!~LawrenceG@S0106001c10b7770f.pk.shawcable.net] has joined #linuxcnc-devel
[21:26:39] -!- bradsimantel has quit [Ping timeout: 276 seconds]
[21:32:00] -!- kwallace [kwallace!~kwallace@tmb-240.sonnet.com] has joined #linuxcnc-devel
[21:35:59] <PCW> andypugh: how did you set the nvencodermodeN ? is there code in master for this? (we would like to check the whole process)
[21:39:09] <andypugh> I use setsserial
[21:39:21] <andypugh> (man setsserial for vague and confusing instructions)
[21:39:25] <andypugh> It's in Master
[21:39:39] -!- tjb1 has quit [Read error: Connection reset by peer]
[21:41:11] <PCW> does it use or list symbolic names?
[21:43:14] <andypugh> I am not sure I understand the question?
[21:43:41] <andypugh> setsserial cmd="set hm2_5i23.0.7i73.0.1.nvencmode3 5"
[21:44:24] <andypugh> Actually loadrt setsserial... because it's a really strange component :-)
[21:46:19] <PCW> setting the volatile parameters in HAL might be nicer in some ways (state is all visible in HAL rather then hidden in EEPROM on card)
[21:46:21] <PCW> in any case here's a updated 7I73 firmware image: freeby.mesanet.com/7i73.bin
[21:49:56] -!- c60 has quit [Quit: leaving]
[21:57:22] -!- DJ9DJ has quit [Quit: bye]
[22:00:10] -!- mk0 has quit [Quit: Leaving]
[22:07:31] -!- bradsimantel has quit [Read error: No route to host]
[22:10:34] -!- bradsimantel has quit [Client Quit]
[22:12:51] -!- FinboySlick has quit [Quit: Leaving.]
[22:12:58] -!- kmiyashiro has quit [Quit: kmiyashiro]
[22:24:26] -!- Brandonian has quit [Quit: Brandonian]
[22:24:43] -!- vladimirek has quit [Remote host closed the connection]
[22:37:47] -!- sumpfralle has quit [Ping timeout: 246 seconds]
[23:01:38] -!- spiderdi1on has quit [Read error: Operation timed out]
[23:30:31] -!- dedis10 [dedis10!~dedis10@akw403.cs.yale.edu] has joined #linuxcnc-devel
[23:33:28] <KGB-linuxcnc> 03andy 05master d9d985c 06emc2 10src/hal/drivers/mesa-hostmot2/sserial.c * Make sserial a little more patient during startup
[23:34:52] -!- mhaberler [mhaberler!~mhaberler@extern-181.stiwoll.mah.priv.at] has joined #linuxcnc-devel
[23:41:04] <PCW> I wonder if there was a bug in Micges's code so it just endlessly XMITed or checked the recv packet rather that the requires xmit/recv per poll
[23:46:08] <andypugh> I don't know, but the wait is a lot more elegant.
[23:47:00] <PCW> elegant?
[23:47:31] <andypugh> Handing some CPU time back, rather than frantically polling the card.
[23:48:24] <PCW> OK that good because it could take a while to timeout if there's a fault
[23:49:46] <PCW> but normal response will be less than 100 uSec or so, so you dont want to wait too long
[23:50:46] <andypugh> It waits in blocks of 50uS