Guest847 is now known as skunkworks
wow -- xilinx webpack is now up to a 1.6GB download
heh - yes indeed. it's a biggie
and what's really nice is that you need to download that, tehn unpack it, and then you can install (which decompresses it)
so a system with a small hard drive requires special care :)
how often do they version bump that xilinx stuff? seems like it will be pain to keep up to date
i have just compiled kernel, rtai and emc2 trunk, now I'm going to analize the code, any hint or guide to look before getting dirt my hands?
jmkasunich_: indeed. apparently the last step of installation is to download an additional few hundred megabytes of "service pack". Apparently that adds up to 284MB + 620MB more ..
eugenio: that's too broad a question to really be answerable. If you have specific questions, including "where's the source code that does X", those are much easier to answer.
there's directory.map at the top level, and some musings in docs/src/code but both probably contain outdated information in some instances
use scripts/swish to do a full-text search of the source and documentation trees
if you use a tags aware editor, use "make tags" in the source directory to make the tags file
the info in directory.map was what i was looking for
i am with xilinx ise 7.1, the biggest problem i have is its shitty driver of the jtag for linux
[Global Notice] Hi all, we're experiencing some routing problems between our US hub and the machine services reside on, I'm about to do some re-routing and will also move services. There will be NO services while we swap things around. Thank you for using freenode and have a great day!
SWPadnos_ is now known as SWPadnos
* jepler is compiling the pluto_servo verilog in webpack for the 7i43 as a way to get familiar with ise
get familiar with the "clean project files" menu item
a few source code changes were required, and of course to really run this I'd have to get the pin constraints right
or, if you find out how to turn off that stinking buggy incremental compile feature, please tell me how to do it
was there a UCF file in the 7i43 zip file? (if there is a 7i43 zip file)
is that why on an earlier run it kept saying one stage of my project had encountered an error, and wouldn't proceed?
yes there are some ucf files in there
I don't think so
and yes, the 7i43 zip file is now available
by the way, I'm serious about sending you one. if you want to play around, I think it's a good investment
I dunno if it would ever make it to the top of my list of things to do
no biggie. I know how those lists tend to be like mobius strips anyway ;)
if you're feeling generous I wouldn't turn you down -- but I don't want you to feel you're bargaining for some kind of software to be written for it
it's not a "deal", it's a "gift, with a possible hope of something good coming out of it, whether or not it's good for me"
of course, I haven't ordered any yet, and I'm busy these days, so don't expect to get one on Monday :)
by christmas, maybe
you never know
of course, I'm going to be on the road basically from next Tuesday until the end of the month, then I may be on the road again for a couple of weeks, then we go off for some Christmas trip, and then .......
anyway, that's very generous of you
heh. I hope it is at least
I think my company will have some money to spend before the end of the year, so I may get a 5-pack or something.
it's nice to be able to deduct expenses for that kind of toy^Wequipment
I am a bit surprised that the speed difference between the acex and the spartan3 is not greater -- the pluto_servo design only gets about 2.2x the clock speed (assuming I correctly remember the timings under altera's software)
.. about 75MHz vs about 180MHz
it's possible that it's limited by the design itself
I think optimization of an FPGA isn't intuitive to a "sequential programmer" - ie us software guys
I wasn't saying my design was or wasn't inherently good, but that I figured fpga technology was a "rising tide that floats all boats"
conceptually, an FPGA doesn't reuse any "code", so you may as well make everything run in parallel and just select which output you want to use at the end
it did float by a factor of 2.2 :)
and I notice that the mesanet samples use the high speed grade
I chose the low speed grade
I wonder how much difference that makes
could be another 50% actually
well, 1.5x the 2.2x, so 3.3x
the speed grades are -4 and -5 but I dunno what that means
I think the low speed ones are 6ns, adn the high are 4ns - or it's in that ballpark anyway
that may be ns, but I'm not sure it is
hmmm. no, it probably isn't, because I think there are some -1s
SWPadnos: does the free version of ise benefit from multiple CPUs?
(or multiple cores, to be more accurate)
jepler, it didn't look like it, but I may have missed a configuration setting or two
03jepler 07v2_2_branch * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: restore cone-rotating behavior (danielbr misses it)
uh oh, I just did something very different from what I intended
03tissf 07TRUNK * 10emc2/src/po/ (fr.po fr_axis.po fr_rs274_err.po): French translation update
% cvs up -j1.31 -j22.214.171.124 axis.py
cvs commit -m"merge from v2_2_branch: restore cone-rotating behavior" axis.py
new revision: delete; previous revision: 1.131
now, how do I undo this little screwup I just made?
cradek, help! :)
03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot2_log.txt
yeah I bet the compile failed!
03compile-farm 07BDI-4.51 (126.96.36.199-rtai) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot6_log.txt
03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: undo incorrect merge that deleted this file
03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot1_log.txt
03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: merge from v2_2_branch: restore cone-rotating behavior
tissf: hi tissf
03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build PASSED
03compile-farm 07BDI-4.51 (188.8.131.52-rtai) * 10emc2head/: build PASSED
damn, that's a lot faster now
yeah it is
tissf: the translation updates you made, should they also be made in 2.2.x?
tissf: if so, I can help you learn about using branches in cvs
Yes i think
Just finish a fresh dapper install and emc2.trunk compilation (2 days ! )
tissf: CVS uses branches so that different versions of the software can be maintained. So far, you have always been working with the one called the TRUNK.
tissf: but since the release of emc 2.2.0, there is another branch called v2_2_branch. On that branch, we make only a few kinds of changes: documentation improvements, translation improvements, bug fixes, and new drivers
03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build PASSED
jepler: ok I have to work on v2_2_branch ?
tissf: to make changes in v2_2_branch, you first have to get a CVS checkout of the branch. Do that by adding "-rv2_2_branch" to your "cvs co" command, and also give the checkout a distinctive name with the "-d<dirname>" flag
in that directory, you can use "cvs up" and "cvs commit" like you've already learned
in most cases you'll first make, test and check in your changes in the TRUNK. then you'll make the same changes in v2_2_branch, test, and check in a second time
CVS can help you make the same changes in the branch as on the TRUNK. If you noted the old and new version numbers when you committed a file on TRUNK, you can use a special form of "cvs up" to make the same changes in the file on the branch: cvs up -jOLDVERSION -jNEWVERSION filename
you can use "cvs log" or cvsweb (http://cvs.linuxcnc.org/cvs/emc2/)
to help you find the revision numbers if you didn't note them when you made the first check-in
03jepler 07v2_2_branch * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: actually restore A-axis rotation of cone
03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: merge from trunk: actually restore A-axis cone rotation
tissf: after the special "cvs up", you still have to "cvs commit"
so for instance in the case of fr_axis.po, you would run this command in the src/po directory: cvs up -j1.12 -j1.13 fr_axis.po
then test and commit
jepler: how test ?
tissf: build the software and verify that the new translations appear
jepler: will do everything
good, thank you
jepler: thank you for help
jepler: the cvs up in the src/po directory but which version? trunk or v2_2_
tissf: in the v2_2_branch directory -- the one that you want to modify
jepler: ok thanks
that "cvs up" just changes the file on your local disk, it doesn't change anything on the cvs server until you commit
jepler: Silly question sorry!
SWPadnos: seems -5 is the high speed grade and -4 the low one! that's the opposite of what I expected.
ok, I guess that proves the numbers aren't nanoseconds :)
python ../docs/src/lyxtree.py --imagedir=../docs/html -o objects/gui_axis.xml ../docs/src/gui/axis.lyx
# Using new toc
Traceback (most recent call last):
File "../docs/src/lyxtree.py", line 729, in ?
doc, outfile = parse(sys.argv[1:])
File "../docs/src/lyxtree.py", line 720, in parse
File "../docs/src/lyxtree.py", line 642, in IndexFixer
NameError: global name 'term' is not defined
make: *** [objects/gui_axis.xml] Erreur 1
hm I must have screwed it up
compile error on the v2_2_branch
noerror with trunk
update and try again?
03jepler 07v2_2_branch * 10emc2/docs/src/lyxtree.py: I am incompetent today