Back
[11:51:06] <skunkworks> pcw_home, any thing I can do to trouble shoot this?
[11:51:38] <skunkworks> I would love to have an image of your install.. Do you have a 7i80 you could test?
[11:52:39] <pcw_home> I have a 7I76E here (its identical Ethernet wise)
[11:54:40] <skunkworks> says you! ;)
[11:55:07] <skunkworks> I reduced the memory to <1gb as a test. I am grasping at straws
[11:55:35] <skunkworks> it has acted the same in multible systems so I think I am doing something stupid - just don't know what
[11:58:45] <skunkworks> the only thing different I see on yours is the ip address is different.. have you tried the default rom ip (192.168.1.121)
[12:15:51] <pcw_home> that will screw stuff up if you have another Ethernet/wireless in the same range
[12:16:42] <pcw_home> but if all other adapters are disabled it should make no difference at all
[12:17:06] <skunkworks> sure - I have only 1 nic (onboard) with this setup.. The atom board
[12:17:19] <Tom_itx> no mb nic?
[12:18:26] <skunkworks> it is a mb nic
[12:23:36] <pcw_home> Really no idea (tried a 7I76E and 7I80 at work with the G945, Asus, Intel D41, an ASUS AMD 450E not one hiccup)
[12:24:48] <skunkworks> darn
[12:27:52] <skunkworks> you don't think you did any extra than the directions you gave me?
[12:32:03] <pcw_home> I dont think so, I installed on the HD I tested at at work with the script that I showed you
[12:40:30] <pcw_home> http://imagebin.org/284751
[12:40:31] <pcw_home> 7I76e running at 1 KHz and being pinged at the same time...
[12:45:40] <skunkworks> neat! and crap. :)
[12:46:03] <skunkworks> so - other than the bit file - is there a firmware that could be different in the 7i80?
[12:46:09] <skunkworks> or is the bit file it?
[12:46:28] <eric_unterhausen> it's all in the wrist
[12:47:28] <pcw_home> Nope all 7I80/7I76E firmware is identical as far as the Ethernet goes (only the hostmot2 stuff may be different)
[12:49:25] <pcw_home> Card specific stuff is in Ethernet EEPROM (IP address, LED options etc)
[12:50:13] <pcw_home> card name,mac address
[12:58:51] <skunkworks> I just loaded your config. I got a
[12:59:01] <skunkworks> encoder 0 : quadautre count error
[12:59:21] <skunkworks> (only spelled correclty_
[13:01:01] <pcw_home> not meaningful
[13:01:07] <pcw_home> (a bug in master)
[13:01:10] <skunkworks> ah
[13:03:33] <skunkworks> with your config - the second I set the watchdog has bit to false - I get the encoder error and it bits again.
[13:03:45] <skunkworks> wth - I will try another board
[13:11:06] <pcw_home> do you have a direct MAC --> 7I80 connection? (a switch would cause troubles)
[13:13:08] <skunkworks> direct connection
[13:15:41] <skunkworks> if I run a hal file with just pet, read write. the watchdog bites.
[13:17:22] <pcw_home> beware minimal hal files (all I/O is enable if not mentioned)
[13:18:04] <pcw_home> enabled
[13:26:11] <pcw_home> but your 8 ms ping time indicates a problem unrelated to anything higher up
[13:28:00] <skunkworks> this computer - the ping times stay less than 300us
[13:28:17] <skunkworks> normally <100
[13:28:57] <skunkworks> the watchdog bite seems to only last as long as the timeout is set. nothing more
[13:29:39] <skunkworks> 10000000 should be 10 seconds?
[13:29:48] <skunkworks> 100000000000
[13:29:50] <skunkworks> I mean
[13:30:04] <skunkworks> heh
[13:30:15] <skunkworks> this computer - the ping times stay less than 300us
[13:30:20] <skunkworks> normally <100
[13:30:36] <skunkworks> 100000000000 should be 10 seconds?
[13:31:55] <pcw_home> thats 100 (which will wrap so unknown)
[13:32:54] <skunkworks> oh
[13:34:34] <pcw_home> mine is set for 10 ms
[13:39:25] <skunkworks> If I change the mac address - I get the same result..
[13:41:08] <pcw_home> The mac address will only make a small difference (avoiding ARP queries every couple minutes)
[13:41:15] <skunkworks> ok
[13:44:11] <pcw_home> Rather than debug things higher up, I would
[13:44:13] <pcw_home> sudo ping -i .001 [7i80IPADDR] >pin.txt
[13:44:16] <pcw_home> for a couple minutes and see if you have any large times
[13:51:06] <skunkworks> http://pastebin.ca/2520934
[13:51:15] <skunkworks> .376?
[13:51:56] <pcw_home> Thats about what I have here
[13:52:27] <skunkworks> ok good
[13:52:55] <skunkworks> wonder what the difference is
[13:54:05] <pcw_home> Pretty mysterious (been running here all morning and several days at work on a D525)
[13:58:07] <skunkworks> when was the last time you built? are you running
[13:58:07] <skunkworks> ^Cempire@empire-OptiPlex-380:~/linuxcnc-dev/ubc3-7i80$ git describe
[13:58:07] <skunkworks> v2.5.3-3608-g60e91dc
[13:59:34] <pcw_home> pcw@pcw-G41M-Combo:~/linuxcnc-dev/ubc3-7i80$ git describev2.5.3-3608-g60e91dc
[13:59:46] <skunkworks> crap crap!
[14:11:18] <skunkworks> pcw_home, any help?
http://pastebin.ca/2520936
[14:13:11] <skunkworks> happy new year btw
[14:18:56] <pcw_home> Happy new year
[14:19:21] <pcw_home> Hmm I dont have such a log maybe thats the problem...
[14:20:21] <skunkworks> you have to do a sudo make log in the src directory
[14:20:27] <skunkworks> (I just did it )
[14:20:51] <skunkworks> then the log is in the /var/log/linuxcnc.log
[14:29:54] <pcw_home> that fails for me
[14:30:24] <skunkworks> http://static.mah.priv.at/public/html/common/UnifiedBuild.html#_installation
[14:33:38] <skunkworks> wonder why that would fail
[14:38:56] <pcw_home> no idea:
[14:39:04] <pcw_home> make: Entering directory `/home/pcw/linuxcnc-dev/ubc3-7i80/src'
[14:39:04] <pcw_home> *** configure.in has changed ***
[14:39:04] <pcw_home> Not running autoconf to avoid tainting universal build
[14:39:04] <pcw_home> Please investigate or start from scratch:
[14:39:04] <pcw_home> ./autogen.sh && ./configure --your-configure-args...
[14:39:05] <pcw_home> make: *** [configure] Error 1
[14:39:05] <pcw_home> make: Leaving directory `/home/pcw/linuxcnc-dev/ubc3-7i80/src'
[14:45:31] <skunkworks> did you try ./autogen.sh and your ./configure.....
[14:47:52] <skunkworks> this is running DEBUG=5
[14:48:04] <skunkworks> http://pastebin.ca/2520944
[15:00:24] <pcw_home> 7I76E:
[15:00:26] <pcw_home> http://pastebin.ca/2520945
[15:04:08] <micges> skunkworks: do you have sserial enabled?
[15:04:32] <skunkworks> I don't think so
[15:05:07] <skunkworks> sserial_port_0=00000000
[15:05:22] <skunkworks> hm2/hm2_7i80.0: parsing config string "num_encoders=1 num_pwmgens=1 num_stepgens=1"
[15:05:52] <micges> yes I'm looking now
[15:07:38] <skunkworks> looks like it created the pins... But I have the same issue with peters config
[15:09:21] <skunkworks> (sorry - it didn't create the pins - I was looking at peters log_
[15:15:41] <skunkworks> THis is running peters config
[15:15:42] <skunkworks> http://pastebin.ca/2520949
[15:17:43] <skunkworks> this is running the config with DEBUG=5
[15:17:43] <skunkworks> http://pastebin.ca/2520950
[16:00:26] <skunkworks> micges, any thoughts?
[16:00:49] <micges> not really
[16:01:30] <micges> lets summarize:
[16:01:36] <micges> same mb?
[16:02:01] <skunkworks> I have tried 4 different systems - 2 different installs of ubuntu.
[16:02:14] <skunkworks> one of them is the same as pcw test systems.
[16:02:40] <micges> and on all you have wd bite?
[16:02:47] <skunkworks> yes
[16:03:26] <skunkworks> this system has the network manager uninstalled and I do a
[16:03:35] <skunkworks> sudo ifconfig eth0 192.168.1.1
[16:03:46] <skunkworks> on startup
[16:05:49] <pcw_home> That is a difference, I didn't change anything in the network setup other than using the ubuntu tools
[16:06:01] <skunkworks> I tried that also first..
[16:06:16] <skunkworks> I removed the network manager as a test
[16:07:01] <micges> skunkworks: try short run with debug=1 and pstebin output
[16:07:12] <pcw_home> (the network mangler is dangerous since if you fuss with the tool when linuxcnc is up it causes trouble)
[16:07:13] <micges> in loadrt hm2_eth
[16:15:06] <skunkworks> http://pastebin.ca/2520957
[16:16:22] <micges> pcw_home: do you see sth suspicious on wd at line >200 ?
[16:21:55] <pcw_home> might be nice to know the data
[16:24:33] <micges> skunkworks: can you make little code change for tests?
[16:24:45] <skunkworks> sure
[16:25:12] <micges> hm2_eth.c
[16:26:47] <pcw_home> write cookie
[16:26:50] <pcw_home> read status
[16:26:50] <pcw_home> write timeout
[16:26:52] <pcw_home> clear WD
[16:26:54] <pcw_home> write cookie
[16:26:54] <pcw_home> read status
[16:26:56] <pcw_home> (WD has bit!)
[16:27:40] <pcw_home> is how I decode that section
[16:27:51] <pcw_home> need better time stamps though
[16:28:33] <micges> skunkworks: sorry tram.c
[16:28:45] <micges> line 172
[16:29:04] <micges> change 'queue_write' to 'write'
[16:30:58] <skunkworks> then make - and run - post a log?
[16:32:41] <micges> yes
[16:33:19] <micges> still with debug=1 option
[16:37:18] <skunkworks> http://pastebin.ca/2520958
[16:39:49] <micges> ok, reverse that change
[16:39:54] <micges> go to hm2_eth.c
[16:40:05] <micges> line 275
[16:42:43] <micges> change whole line 275 and 276 to:
[16:42:45] <micges> LL_PRINT_IF(debug, "write(): PACKET SENDED [CMD:%02X%02X | ADDR: %02X%02X | DATA: %02X%02X%02X%02X | SIZE: %d]\n", packet.wr_packet.cmd_hi, packet.wr_packet.cmd_lo,
[16:42:45] <micges> packet.wr_packet.addr_lo, packet.wr_packet.addr_hi, packet.tmp_buffer[0], packet.tmp_buffer[1], packet.tmp_buffer[2], packet.tmp_buffer[3], size);
[16:43:22] <micges> make, run and pastebin log
[16:43:45] <skunkworks> can you pastebin that line?
[16:43:55] <micges> sure
[16:44:43] <micges> http://pastebin.ca/2520964
[16:52:34] <micges> skunkworks: I didn't compile it
[16:53:33] <pcw_home> the grammar nazi in me suggests PACKET SENT instead of PACKET SENDED
[16:54:36] <micges> thanks
[16:55:15] <skunkworks> http://pastebin.ca/2520965
[16:55:23] <skunkworks> does that make sense?
[17:00:11] <micges> pcw_home: ?
[17:00:34] <micges> pcw_home: for me it's ok
[17:01:18] <pcw_home> Looks OK to me
[17:02:15] <pcw_home> (other than the WD biting immediately even thou it has a very long timeout: 0x1FA10700
[17:02:44] <pcw_home> maybe the read data is bogus
[17:03:56] <skunkworks> swapped different network cable also
[17:04:51] <pcw_home> well bad data would not get this far
[17:05:33] <pcw_home> where is the debug=1 flag (I will duplicate here)
[17:05:49] <micges> in loadrt hm2_eth
[17:06:47] <skunkworks> would seeing the read data help?
[17:07:41] <micges> prpbably
[17:11:04] <micges> skunkworks: add this line:
http://pastebin.ca/2520967
[17:12:03] <micges> after line 230
[17:16:51] <pcw_home> I cant see that my config sets the watchdog timeout
[17:17:14] <skunkworks> http://pastebin.ca/2520968
[17:17:39] <skunkworks> you have the watchdog timout set in the hal file
[17:18:04] <skunkworks> or do you mean in the log?
[17:18:22] <skunkworks> READED DATA: FFFFFF00
[17:19:17] <pcw_home> that is not a bite!
[17:19:50] <pcw_home> bite is 1 in LSb
[17:20:46] <skunkworks> something is goofy ;)
[17:21:33] <micges> if (hm2->watchdog.status_reg[0] & 0x1) {
[17:21:33] <micges> 108 HM2_PRINT("Watchdog has bit! (set the .has-bit pin to False to resume)\n");
[17:21:33] <micges> 109 *hm2->watchdog.instance[0].hal.pin.has_bit = 1;
[17:21:33] <micges> 110 hm2->llio->needs_reset = 1;
[17:21:33] <micges> 111 }
[17:23:05] <pcw_home> also FFFFFF00 is not possible from that register
[17:23:28] <pcw_home> only 0x00000000 or 0x00000001
[17:23:51] <pcw_home> if readstatus = '1' then
[17:23:52] <pcw_home> obus(0) <= wdstatus;
[17:23:53] <pcw_home> obus(buswidth -1 downto 1) <= (others => '0');
[17:23:55] <pcw_home> end if;
[17:27:24] * skunkworks stares blindly..
[17:27:37] <micges> skunkworks: this is whole hm2_eth.c to test:
http://pastebin.ca/2520971
[17:32:38] <skunkworks> micges,
http://pastebin.ca/2520973
[17:33:29] <micges> hmm correct data size readed
[17:34:13] <pcw_home> if i read 0x0d00 I get 0x00000001
[17:35:52] <micges> even so wd code only checks one lsb, why is it triggering?
[17:38:00] <skunkworks> so what would be different between mine and peters..
[17:38:22] <skunkworks> and is that the packet that is triggering the watchdog?
[17:38:26] <pcw_home> I never get the watchdog enabled
[17:40:27] <pcw_home> lots of questions
[17:40:28] <pcw_home> Why dont I have my watchdog enabled (no timeout was written)
[17:40:30] <pcw_home> How do you get 0XFFFFFF00 (its not a possible response from a read of 0x0D00)
[17:40:31] <pcw_home> How does 0xFFFFFF00 masked with 1 return true
[17:41:14] <skunkworks> so - yours works because the watch dog isn't enabled?
[17:41:18] <micges> if so then I have it disabled too...
[17:42:29] <pcw_home> skunkworks did you try my config as-is?
[17:44:12] <pcw_home> its just hm2-stepper. the watchdog should be enabled but appears not to be (I dont see any writes to 0x0C00 in the log)
[17:44:14] <pcw_home> but maybe they have been lost somehow
[17:44:43] <micges> why can I read encoder and can't read wd?
[17:45:05] <micges> if so it seems writing must be checked deeply
[17:45:48] <skunkworks> pcw_home, I had to change the ip, mac and board.. I don't know if I changed anything else
[17:45:57] <pcw_home> Sorry it does look like my watchdog is enabled
[17:46:10] <pcw_home> >0142000d
[17:46:12] <pcw_home> > 00000000
[17:46:15] <pcw_home> >0142000c
[17:46:15] <pcw_home> > 2f3f0f00
[17:46:24] <skunkworks> If I don't add the watchdog function - I can't really jog around - I get following errors
[17:46:54] <skunkworks> (and then set the watchdog has bit to true
[17:48:09] <skunkworks> I am going to have to leave..
[17:48:33] <skunkworks> I will be around to morrow
[17:48:39] <skunkworks> tomorrow
[17:49:49] <skunkworks> Thank for helping with this. I have no clue what is going on.
[17:51:01] <pcw_home> Yeah my watchdog is enabled (the timeout is set)
[17:51:34] <pcw_home> Thank as well, a pretty subtle bug
[17:51:41] <pcw_home> thanks
[17:52:11] <skunkworks> I just have a hal file that only runs the 3 threads.. nothing else. PRetty basic
[17:52:26] <skunkworks> (watchdog, read and write)
[17:52:46] <skunkworks> BBL
[17:52:54] <pcw_home> bye
[17:54:39] <pcw_home> btw micges I can read hm2 registers from a terminal while linuxcnc is running which is pretty helpful
[17:54:41] <pcw_home> (I can see that the watchdog is enabled and working on my config this way)
[17:55:11] <micges> by mf?
[17:55:48] <micges> or by simple python script
[17:59:04] <micges> (hmm by both really)
[18:03:38] <pcw_home> i just used python but anything should work
[18:04:49] <pcw_home> watchdog timer value with linuxcnc running and stopped
[18:04:50] <pcw_home> >0142000c
[18:04:52] <pcw_home> > b5330f00
[18:04:53] <pcw_home> >0142000c
[18:04:55] <pcw_home> > ffffffff
[18:05:25] <micges> correct
[18:17:16] <pcw_home> >01420011
[18:17:17] <pcw_home> > 3f040000
[18:17:19] <pcw_home> >01420011
[18:17:21] <pcw_home> > 00000000
[18:19:17] <micges> wrong?
[18:19:42] <pcw_home> (DDR register linuxcnc running and stopped)
[18:19:46] <pcw_home> I stopped linuxcnc by pulling the Ethernet cable
[18:19:48] <pcw_home> so it did not clean up so this indicates that the WD
[18:19:48] <pcw_home> is working and held of by linuxcnc and bites when left alone
[18:19:59] <pcw_home> all working as expected
[18:20:17] <micges> ok
[18:20:31] <pcw_home> s/held of/held off/
[18:21:25] <pcw_home> whats different between my setup and skunkworks is a mystery
[18:22:28] <pcw_home> Andy has it (preempt_rt/7I80) working also (did you see his washing machine motor video?)
[18:22:52] <micges> yes
[19:31:08] <skunkworks> logger[mah]:
[19:31:08] <logger[mah]> skunkworks: Log stored at
http://linuxcnc.mah.priv.at/irc/%23linuxcnc-devel/2014-01-02.html
[19:41:26] <skunkworks> could it be something that I pulled for dependancies that is different than yours?
[19:45:02] <pcw_home> Maybe (I did find that my watchdog is enabled and working fine)
[19:45:36] <skunkworks> I saw that
[19:46:38] <pcw_home> if you have a moment, can you read the wd status register?
[19:46:59] <pcw_home> I think mesaflash can do it
[19:49:32] <skunkworks> sorry - no.. it isn't here.
[19:49:36] <skunkworks> I am at home
[20:09:01] <Tom_itx> can you flash over parport?
[20:09:16] <Tom_itx> or do you need jtag
[20:10:48] <pcw_home> flash what?
[20:12:23] <Tom_itx> 7i43
[20:12:41] <Tom_itx> not that i'm gonna, i wanna know how
[20:13:11] <Tom_itx> i see a few utilities in the mesa directory
[20:13:20] <pcw_home> Yes you can flash the 7I43 via EPP
[20:13:33] <pcw_home> or USB if your 7I43 has it
[20:13:43] <Tom_itx> what sort of file do you upload to it?
[20:13:47] <Tom_itx> yes it has usb
[20:14:25] <Tom_itx> i don't see mesaflash in the 7i43 dir though
[20:15:31] <pcw_home> I'm not sure mesaflash supports EPP flashing anything but 7I90s now (but the 7I43 should be easy to add)
[20:15:58] <pcw_home> if you want to, current is probably easier to use USB and windows
[20:16:07] <pcw_home> currently
[20:16:28] <Tom_itx> is it the same utility as found in the 5i25 directories?
[20:16:40] <pcw_home> in all cases its a bootstrap process
[20:16:59] <pcw_home> yes same utility for PCI/PCIE/Ethernet/EPP
[20:18:10] <Tom_itx> easier done in windows or linux?
[20:18:20] <pcw_home> (maybe serial as well, currently trying to finish serhm2 for a customer using 7I90s)
[20:18:35] <pcw_home> and SPI...
[20:19:02] <Tom_itx> what is mflash.exe in the dos directory?
[20:19:03] <pcw_home> easier done in windows since the utility is there
[20:19:20] <pcw_home> for the 7I43
[20:20:07] <pcw_home> mflash will do it as well (under DOS) or the older SC7I43P and SCM7I43P
[20:20:12] <Tom_itx> are there docs explaining mesaflash?
[20:20:37] <pcw_home> it has some help: mesaflash --help
[20:20:42] <Tom_itx> ok
[20:24:47] <Tom_itx> where would i find the latest firmware file for it?
[20:25:19] <pcw_home> firmware?
[20:25:22] <Tom_itx> i43cpld5.jed?
[20:25:43] <Tom_itx> err 4
[20:26:03] <pcw_home> thats the CPLD "firmware"
[20:26:17] <Tom_itx> that's not what i want to update?
[20:26:24] <pcw_home> that _does_ require JTAG
[20:26:31] <Tom_itx> i figured it probably did
[20:26:40] <pcw_home> no you just use normal bitfiles
[20:26:42] <Tom_itx> i may have an old jtag that would work
[20:26:50] <Tom_itx> on an old dev board for xilinx i have
[20:27:43] <Tom_itx> i better just leave it. i'd probably just screw it up
[20:27:57] <pcw_home> I dont understand what you are trying to do
[20:28:29] <Tom_itx> i'm just poking around a bit
[20:28:38] <Tom_itx> nothing really required
[20:28:49] <Tom_itx> just trying to learn a bit more about them
[20:29:26] <pcw_home> if you move your 7I43 jumpers for flash config I think you will find that it autoconfigs
[20:29:42] <Tom_itx> one thing might be a newer sserial driver
[20:29:56] <Tom_itx> firmware
[20:30:08] <Tom_itx> not that it's needed really
[20:30:15] <Tom_itx> just trying to learn the process
[20:30:22] <pcw_home> (theres _some kind of bitfile_ installed for test)
[20:31:18] <pcw_home> ver 43 is the latest released sserial host firmware
[20:31:41] <Tom_itx> i don't remember what mine was, 41 maybe
[20:31:59] <Tom_itx> i'd have to check for sure
[20:32:20] <pcw_home> the source in the zip has 43
[20:33:05] <pcw_home> you would need to remove sslbprom41 and add sslbprom43 from your source list to update
[20:33:33] <pcw_home> everything after 40 has been pretty minor
[20:34:27] <Tom_itx> where are those located?
[20:35:51] <pcw_home> in the source zip file
[20:36:08] <pcw_home> sslbprom43.vhd is the latest
[20:36:21] <Tom_itx> i'll dig a bit deeper
[20:37:01] <Tom_itx> i must not have it unzipped yet
[20:37:12] <skunkworks> pcw_home: does the mesaflash work for 12.04 or do I need to build it from source?
[20:37:53] <pcw_home> i just built it
[20:38:04] <pcw_home> takes about 1 second
[20:38:12] <Tom_itx> me or skunkworks?
[20:38:14] <skunkworks> ok - no suprises?
[20:38:31] <skunkworks> I remember having to edit the source the last time I tried.
[20:38:39] <pcw_home> let me dig up micges sources
[20:38:53] <pcw_home> I had to type make
[20:39:01] <skunkworks> heh
[20:40:19] <pcw_home> freeby.mesaflash.com/mesaflash3.zip is the latest I have
[20:42:11] <pcw_home> pcw@pcw-G41M-Combo:~/mesaflash$ ./mesaflash --ip=10.10.10.10 --device 7i80 --rpo 0x0d00
[20:42:11] <pcw_home> 0x00000001
[20:43:08] <pcw_home> (watchdog status)
[20:45:04] <Tom_itx> ahh found it
[20:45:54] <Tom_itx> does the sslbprom43.vhd turn into a bit file then you upload that via mesaflash?
[20:46:07] <Tom_itx> via usb or pport
[20:46:18] <pcw_home> Tom_itx: as an interesting exercise you could port HostMot2 to your xilinx dev board
[20:46:29] <Tom_itx> not big enough
[20:46:32] <Tom_itx> it's a tiny chip
[20:46:40] <Tom_itx> or i might try it
[20:46:43] <pcw_home> make a tiny config
[20:46:50] <skunkworks> pcw_home: I should run that while the hal file is running?
[20:47:17] <pcw_home> should be OK
[20:47:39] <pcw_home> let me try while linuxcnc is running
[20:47:40] <Tom_itx> the sslbprom43 is firmware for the cpld?
[20:47:49] <Tom_itx> where does the hostmot2 stuff load then?
[20:48:15] <pcw_home> sslbprom is just code for an embedded CPU that runs the host part of sserial
[20:49:05] <pcw_home> its easy to get confuse when there is firmware running on the firmware
[20:49:16] <Tom_itx> no kidding
[20:49:22] <Tom_itx> i sorta get it
[20:49:52] <Tom_itx> what is seveni43.xise for?
[20:50:09] <pcw_home> thats the project file for 7I43
[20:50:22] <Tom_itx> does it pull in all ^^ those files?
[20:50:31] <Tom_itx> the sserial and hostmot2 stuff
[20:50:39] <pcw_home> yes (look at it, its text)
[20:50:46] <Tom_itx> k
[20:51:17] <Tom_itx> just a couple files :D
[20:51:42] <pcw_home> the zip file even has the sserial source code but its known to be harmful to look at :-)
[20:52:10] <pcw_home> and the Ethernet source code
[20:52:14] <Tom_itx> is that what andy was messin with?
[20:53:02] <pcw_home> no one sane would mess with anything that low level
[20:55:54] <pcw_home> skunkworks:
[20:55:55] <pcw_home> pcw@pcw-G41M-Combo:~/mesaflash$ ./mesaflash --ip=10.10.10.10 --device 7i80 --rpo 0x0d00
[20:55:57] <pcw_home> 0x00000000
[20:55:58] <pcw_home> (watchdog status with linuxcnc running)
[20:58:46] <skunkworks> ok
[20:59:07] <pcw_home> pcw@pcw-G41M-Combo:~/mesaflash$ ./mesaflash --ip=10.10.10.10 --device 7i80 --rpo 0x0d00
[20:59:08] <pcw_home> 0x00000001
[20:59:10] <pcw_home> (after exiting linuxcnc)
[21:00:18] <skunkworks> so - if it shows 0x000000 while linuxcnc is running - then it isn't comunicating correctly with hal?
[21:00:58] <pcw_home> 0x00000000 is good it means the WD has not bit
[21:02:32] <pcw_home> thats what I was talking about earlier, somehow you had 0xFFFFFF00 data which is impossible
[21:04:18] <pcw_home> so an independent reading of the watchdog status register at 0x0d00 seems in order
[21:11:14] <skunkworks> does the ouput of configure give the versions of the installed dependencies? could that be used to compare what we have?
[21:18:15] <pcw_home> http://pastebin.ca/2521565
[21:20:51] <pcw_home> command line is the one from the script:
[21:20:53] <pcw_home> ./configure --with-rt-preempt --enable-drivers --without-libmodbus
[21:24:32] <skunkworks> yes - that is what I used also - I will compare tomorrow
[21:39:40] <Tom_itx> pcw_home what's the difference between the hm2epp and the hm2epps projects?
[21:41:10] <pcw_home> hm2eep is for 7I43, hm2eeps is for 7I90
[21:41:12] <pcw_home> They could have been combined but there are enough differences to be a nuisance
[21:41:52] <Tom_itx> ok
[21:42:21] <pcw_home> s is standalone (you cannot bootstrap a 7I90 with no FPGA config except by JTAG)
[21:43:19] <pcw_home> (one reason the 7I90 has 2 flash chips so you have to make two mistakes before you brick the card)
[21:46:10] <skunkworks> can you recover them if people send them back?
[21:46:23] <Tom_itx> with jtag probably
[21:49:23] <pcw_home> sure though its very uncommon
[21:50:28] <pcw_home> on spartan6 you also have fallback so if you say pull the PC plug when
[21:50:29] <pcw_home> updating the firmware, the FPGA will boot from its fallback config
[21:50:54] <pcw_home> (true of 7I90 as well)
[21:51:09] <Tom_itx> what happens if you leave the use work.@card@.all enabled as well as the similar card line? does it build all the bit files?
[21:51:31] <pcw_home> Only with the correct batch file
[21:52:33] <pcw_home> (which replaces @card@ and @PIN@ with items from a list)
[21:53:22] <Tom_itx> what file would i need to build if i were to want to update to the 43 firmware?
[21:53:36] <Tom_itx> the hm2epp project or something else?
[21:53:55] <pcw_home> hm2epp or 7i43
[21:54:12] <Tom_itx> i notice it pulls in the sslbprom43 file
[21:54:54] <Tom_itx> how does that get uploaded differently than a regular bit file then?
[21:54:56] <pcw_home> yes it should
[21:55:24] <Tom_itx> and what do i need to change to build that instead of a bit file?
[21:56:47] <pcw_home> you dont build it you just build your bitfile (and if it uses sserial, it will incorporate sslbp43 just like any other source file)
[21:57:00] <Tom_itx> hmm
[21:57:16] <Tom_itx> i didn't check the ver after adding sserial to them
[21:58:16] <pcw_home> if you build with sslbprom43 you will see your sserial version changeto 43 in dmesg
[21:58:28] <Tom_itx> guess i'll have to check that
[21:59:50] <pcw_home> somewhere theres a disassembler for that code...
[22:00:36] <pcw_home> as an aside you can load new sserial firmware live
[22:01:36] <pcw_home> (we do to avoid having to re-compile the FPGA for every code change when developing)
[22:02:41] <Tom_itx> i suppose that takes a while
[22:03:53] <pcw_home> It takes seconds to assemble the sserial source but minutes to re-compile the FPGA bitfile
[22:04:21] <atom1> yeah it's showing ver 43 for sserial
[22:05:07] <atom1> i think it was 41 before i downloaded the new zip
[22:06:20] <pcw_home> sserial is pretty stable now I dont anticipate many changes unless we add features
[22:06:34] <atom1> how full is the fpga with a decent size config?
[22:09:17] <pcw_home> Not sure what a decent size config is but really depends on whats there
[22:09:18] <pcw_home> (resolver, sserial,stepgen,encoder,pwmgen is about the rough size sequence (large to small) )
[22:10:18] <pcw_home> (and of course sserial # of channels affects size as each channel needs a UART)
[22:12:09] <pcw_home> even the 5I20 (smallest HostMot capable FPGA card we make)
[22:12:10] <pcw_home> can have sserial (but not resolver, just too big/too much ROM/RAM)
[22:12:36] <Tom_itx> is the 5i25 larger in size?
[22:13:07] <Tom_itx> i know it was made to be an 'affordable' alternative for alot of systems
[22:15:10] <pcw_home> Yes the FPGA on the 5I25 is bigger then the Spartan3 400K
[22:15:31] <skunkworks> how big is the 7i80-16?
[22:15:45] <pcw_home> bigger still...
[22:15:48] <skunkworks> cool
[22:16:05] <pcw_home> maybe about 4 5I20s
[22:16:20] <skunkworks> if I can get it working - it may do some resolver work
[22:16:57] <skunkworks> *when I get it working
[22:17:04] <Tom_itx> there ya go...
[22:17:14] <pcw_home> I have some work to do before the 7I80 will work with resolvers
[22:18:13] <pcw_home> (its clock low is 100MHZ and the resolver code expects 50 MHz)
[22:18:23] <Tom_itx> do you guys assemble your own boards or farm them out?
[22:18:32] <skunkworks> no rush :)
[22:18:39] <pcw_home> we do no assembly
[22:20:11] <Tom_itx> thanks for all the info, i'm sure you get tired of answering all the silly questions
[22:20:23] <Tom_itx> i'm out for today...
[22:20:26] <pcw_home> not at all
[22:20:30] <pcw_home> 'nite
[22:20:49] <pcw_home> me too
[22:36:45] <JerAgon> Hello linuxcnc developers?...
[22:37:23] <skunkworks> ask your question
[22:38:17] <JerAgon> Have some questions regarding compiling linuxcnc source to an executable
[22:39:00] <JerAgon> Does that make any sense?
[22:39:34] <cradek> ?
[22:40:05] <cradek> what is the problem?
[22:41:49] <cradek> goodnight
[22:43:45] <JerAgon> I have compiled to run in place, but can get it to make as executable?
[23:01:32] <eric_unterhausen> I think the difference is only paths
[23:06:07] <eric_unterhausen> run in place makes an executable file, so I assume that's not really what you wanted to ask
[23:13:31] <JerAgon> I have run through all of the wiki's how to's. They see to very, but this is the just of what I have been doing all with similar results; "all run in place". I would like to be able to compile and have it as packaged. Following are the step I have taken:1. cd linuxcnc-dev/src 2. autogen.sh 3. ./configure # enable run in place 4. make 5. make install-menu 6. sudo make install#, and setuid.
[23:14:23] <JerAgon> Thanks for your help
[23:22:39] <eric_unterhausen> ./configure with no arguments puts everything in the directory structure defined by the git pull
[23:23:01] <eric_unterhausen> --enable-run-in-place is the default
[23:23:34] <eric_unterhausen> looks like ./configure --prefix=prefix is the opposite of run in place
[23:23:53] <eric_unterhausen> I think you have to make install and then make menu-install, that could be wrong though
[23:29:17] <JerAgon> Thanks again! All of this is completly knew to me. I'll try it and let you know.
[23:33:25] <JerAgon> one thing to clarify, would I use the sudo make install?
[23:34:14] <JerAgon> or just make install?