#emc-devel | Logs for 2010-02-14

Back
[01:27:13] <mozmck> LawrenceG: the avr32 chips have audio and usb built in, and sample code as well.
[01:31:50] <LawrenceG> mozmck, thanks... I have a high performance audio codec and I want to stream the data back into the pc.... the generic audio class looks promising... no pc side drivers needed
[01:32:46] <mozmck> general audio class over usb? is your codec on a chip?
[01:37:42] <LawrenceG> yea its a AD1871 chip by Analog devices.. stereo 24 bit 96khz a/d.... I could also write a custom image for a mesa card if I had a few years to learn vhdl :}... the pc is a pain to get data in and out of
[01:38:24] <LawrenceG> in the isa bus days, a protocard and a few gates was all one needed
[01:39:48] <LawrenceG> mind you the pc was so slow, one could not process the data
[01:42:48] <mozmck> yeah, the avr32 is faster than the old pc!
[03:13:49] <skunkworks> cradek: http://cgi.ebay.com/Shaft-alignment-kit-S668CZ-Starrett_W0QQitemZ110483646078QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item19b956a67e
[03:21:56] <jepler> LawrenceG: it looks like an AVR may not be a good choice for your project; I don't think they have clocked serial modes wider than 8 bits, but it looks like the AD1871 really needs one.
[03:33:48] <LawrenceG> yea... 2.3Mb/s is beyond bit banging on most micros...
[03:34:41] <LawrenceG> maybe a meas card interface is more reasonable... A header card for one of the 50 pin solts
[17:13:08] <Guest369> Guest369 is now known as skunkworks
[18:10:59] <alex_joni> all of cia is down
[18:11:20] <alex_joni> but it seems they know about it, and when the datacenter gets back services should get restored.. no ETA so far
[18:15:28] <jt-plasma> is there a deb for testing 2.4?
[18:36:03] <alex_joni> http://emc2-buildbot.colorado.edu/~buildmaster/
[18:36:34] <alex_joni> although the deb line isn't quite sane
[18:37:05] <alex_joni> I think it should be hardy v2.4_branch
[18:37:22] <alex_joni> maybe v2.4_branch-rt ?
[18:37:32] <jepler> the latter, I think
[18:39:49] <jepler> deb http://emc2-buildbot.colorado.edu/~buildmaster hardy v2.4_branch-rt
[18:39:49] <jepler> deb-src http://emc2-buildbot.colorado.edu/~buildmaster hardy v2.4_branch-rt
[18:39:54] <jepler> this is correct; at least, it's installing the package for me
[18:40:48] <jepler> oh, and I put the new hostmot2-firmwares in the 'base' repository on linuxcnc.org
[18:42:28] <cradek> I'm running the v2.4_branch-rt package on jr now, but I didn't figure out the firmwares yet
[18:52:42] <jt-plasma> as soon as I get the Hardinge back on line I'll give them a try
[18:53:10] <jepler> cradek: $ sudo apt-get install hostmot2-firmware-5i20
[18:53:40] <jepler> your firmware should have the same name as before; but others, particularly people with cards that have two fpga sizes available, may differ
[18:54:09] <jepler> bbl
[18:56:23] <cradek> well maybe I do have that installed... I'll check later.
[19:46:48] <jepler> * jepler realizes he could easily use symlinks to make the hostmot2-firmware paths compatible with 2.3
[20:00:03] <Roguish> hey all, back at the hobby after some time off working.
[20:00:57] <Roguish> did a git pull update to emc2-dev, and noticed that there is only a 7i43 directory in /src/hal
[20:01:30] <Roguish> what / where are the other mesa board files? like 5i20?
[21:13:36] <alex_joni> Roguish: it's actually a file
[21:13:46] <alex_joni> hm2_7i43 is for the 7i43
[21:14:06] <alex_joni> hm2_pci is for the pci boards, 5i20, 5i21, 5i22, etc
[21:16:12] <alex_joni> do we care about a temporary replacement for CIA?
[21:20:25] <jepler> alex_joni: not me
[21:24:05] <jepler> Roguish: the hostmot2 firmware images are moved to a new project, called hostmot2-firmware. In the emc2 tree the src/hal/drivers/mesa-hostmot2/firmware/7i43/CPLD directory is the only one left because this part hasn't been moved to hostmot2-firmware. (the CPLD is the chip on the 7i43 board which helps program the FPGA with the appropriate image, and then mediates the EPP communication. Most users never reprogram the CPLD; the so
[21:24:58] <jepler> if you used a symlink in /lib/firmware to your emc2 git checkout to get firmwares for the hostmot2 driver, then you'll have to do something extra now (such as install the new hostmot2-firmware packages through apt-get)
[21:25:58] <jepler> apt-get install hostmot2-firmware-7i43-2 or hostmot2-firmware-7i43-4 depending on the size of FPGA you have (the old S / B in the firmware name), then look in /lib/firmware/hostmot2 for the new firmware name
[21:35:58] <jepler> (in the case of the 5i20 I think all the firmware names are the same..)
[23:30:33] <ries> jepler: your patch about [Emc-developers] [PATCH for v2.3.x] paper over a memory leak could I also apply this on 2.5-dev ?
[23:50:52] <jepler> ries: the code is a bit different in master; I think the patch didn't apply without modification.
[23:51:13] <jepler> the leak exists in master, so some version of the patch should be made there if it tests right
[23:51:15] <ries> jepler: I applied it manually...
[23:51:42] <ries> problem is I get this error during startup of EMC : AttributeError: MyOpengl instance has no attribute 'g'
[23:52:02] <ries> at the line that states : if o.g is not None:
[23:52:40] <jepler> yes, I think the name 'o.g' is different in master, but I don't have the source code in front of me to check. it *might* be called o.canon in master.
[23:53:27] <ries> I will check that...
[23:54:31] <ries> It's indeed o.canon
[23:59:35] <jepler> bbl