03:35
citypw has joined #symbiflow
06:05
proteusguy has joined #symbiflow
08:52
kraiskil has joined #symbiflow
09:17
<
nats` >
litghost, I ran the tiles part in 15 minutes and the nodes part in 55 minutes
09:18
<
nats` >
without tinkering the ratio process blocksize we can do the 074 in 1h10min + few minutes for all file remerge
09:18
<
nats` >
is it a progress ?
09:38
citypw has quit [Ping timeout: 244 seconds]
09:50
kraiskil has quit [Ping timeout: 258 seconds]
10:03
kraiskil has joined #symbiflow
10:09
kraiskil has quit [Ping timeout: 258 seconds]
10:22
kraiskil has joined #symbiflow
10:59
<
kgugala >
@nats` we can discuss 07x fuzzers here, but I have a meeting in about 2 minutes I'll bbl
10:59
<
nats` >
no problem :)
11:00
<
nats` >
I'm at work too
11:37
<
nats` >
I started a run with a full 074 new version script
11:37
<
nats` >
let's see how long it take
12:13
proteusguy has quit [Quit: Leaving]
12:24
kraiskil has quit [Ping timeout: 245 seconds]
12:38
kraiskil has joined #symbiflow
12:45
<
kgugala >
at my side it finished
12:45
<
kgugala >
I'll push the fix
12:47
kraiskil has quit [Ping timeout: 272 seconds]
13:00
kraiskil has joined #symbiflow
13:23
<
nats` >
kgugala, I saw your new bug on the 074
13:23
<
nats` >
shoudl I do something in the new version or is it something solved by ignoring those net ?
13:23
<
nats` >
I didn't fix the working part of the 074
13:23
<
nats` >
the other point to check wouldbe to see if it's because of the 072 patch
13:24
<
nats` >
Command being timed: "python3 run_fuzzer.py"
13:24
<
nats` >
User time (seconds): 14816.12
13:24
<
nats` >
System time (seconds): 755.35
13:24
<
nats` >
Percent of CPU this job got: 346%
13:24
<
nats` >
Elapsed (wall clock) time (h:mm:ss or m:ss): 1:14:53
13:24
<
nats` >
Average shared text size (kbytes): 0
13:24
<
nats` >
Average unshared data size (kbytes): 0
13:24
<
nats` >
Average stack size (kbytes): 0
13:24
<
nats` >
Average total size (kbytes): 0
13:24
<
nats` >
Maximum resident set size (kbytes): 1201000
13:24
<
nats` >
Average resident set size (kbytes): 0
13:24
<
nats` >
Major (requiring I/O) page faults: 0
13:24
<
nats` >
Minor (reclaiming a frame) page faults: 35432521
13:24
<
nats` >
Voluntary context switches: 66357
13:24
<
nats` >
Involuntary context switches: 10470447
13:24
<
nats` >
File system inputs: 11168
13:24
<
nats` >
File system outputs: 27366400
13:24
<
nats` >
Socket messages sent: 0
13:24
<
nats` >
Socket messages received: 0
13:24
<
nats` >
Signals delivered: 0
13:24
<
nats` >
Page size (bytes): 4096
13:24
<
nats` >
Exit status: 0*
13:24
<
nats` >
here is the runtime of the new version of the 074
13:25
<
nats` >
I'm wondering how long it take ont he current version
13:31
<
nats` >
uhhmm kgugala do yo uhave the csv from your zynq run ?
13:31
<
nats` >
I would like to know if it's normal to have a lot of NULL inside it
13:32
<
kgugala >
the bug is for 054
13:32
<
kgugala >
TBH I think that 054 is redundant
13:32
<
kgugala >
it runs the same tcl as 56
13:33
<
nats` >
oh yes sorry
13:33
<
kgugala >
IMO we can drop 54 and use 56 to calculate those
13:33
<
nats` >
I didn't look at 05x serie
13:33
<
kgugala >
I'm testing this right now
13:33
<
nats` >
can you look in the csvoutput of the 074 ?
13:34
<
nats` >
I have a lot of tile NULL
13:34
<
nats` >
but maybe that's ok
13:34
<
nats` >
filetype,subtype,filename
13:34
<
nats` >
tile,NULL,tile_NULL_X0Y156.json5
13:34
<
nats` >
tile,NULL,tile_NULL_X1Y156.json5
13:34
<
nats` >
tile,NULL,tile_NULL_X2Y156.json5
13:34
<
nats` >
tile,NULL,tile_NULL_X3Y156.json5
13:34
<
nats` >
tile,T_TERM_INT,tile_T_TERM_INT_X4Y156.json5
13:34
<
nats` >
tile,T_TERM_INT,tile_T_TERM_INT_X5Y156.json5
13:34
<
kgugala >
I have them too
13:35
<
nats` >
do you have a place to upload it ?
13:35
<
kgugala >
I may have
13:35
<
nats` >
I'm interested in having it to compare with my next run of zynq 074
13:42
<
nats` >
starting the zynq run of my new 074
13:42
<
nats` >
will let you know the time taken :)
13:50
<
nats` >
is there a good to make something like PR on github but only for testing before doing a real PR
13:51
<
kgugala >
nats`, you can always mark a PR as WIP
13:51
<
kgugala >
and use it for discussion
13:52
<
nats` >
ahhh sure !
13:52
<
nats` >
I'll do that for the new version of 074
13:52
<
nats` >
I'll need to have people testing it
14:20
<
nats` >
Work done !
14:20
<
nats` >
Command being timed: "python3 run_fuzzer.py"
14:20
<
nats` >
User time (seconds): 6608.01
14:20
<
nats` >
System time (seconds): 442.71
14:20
<
nats` >
Percent of CPU this job got: 361%
14:20
<
nats` >
Elapsed (wall clock) time (h:mm:ss or m:ss): 32:30.47
14:20
<
nats` >
Average shared text size (kbytes): 0
14:20
<
nats` >
Average unshared data size (kbytes): 0
14:20
<
nats` >
Average stack size (kbytes): 0
14:20
<
nats` >
Average total size (kbytes): 0
14:20
<
nats` >
Maximum resident set size (kbytes): 1020264
14:20
<
nats` >
Average resident set size (kbytes): 0
14:20
<
nats` >
Major (requiring I/O) page faults: 0
14:20
<
nats` >
Minor (reclaiming a frame) page faults: 31386362
14:21
<
nats` >
Voluntary context switches: 37120
14:21
<
nats` >
Involuntary context switches: 6036505
14:21
<
nats` >
File system inputs: 71360
14:21
<
nats` >
File system outputs: 16468960
14:21
<
nats` >
Socket messages sent: 0
14:21
<
nats` >
Socket messages received: 0
14:21
<
nats` >
Signals delivered: 0
14:21
<
nats` >
Page size (bytes): 4096
14:21
<
nats` >
Exit status: 0
14:21
<
nats` >
I'm surprised it went really fast for the zyn
14:24
<
nats` >
same csv file size
14:25
<
nats` >
by eye first and last lines are the same
14:25
<
nats` >
kgugala, are you interested that I push and PR that so you could test it ?
14:38
<
kgugala >
you can tag me in the PR
14:43
proteusguy has joined #symbiflow
15:05
<
nats` >
kgugala, I'm running it from the directory with the makefile to be sure and them I push :)
15:35
<
nats` >
kgugala, I have an other problem with directory in 074 I think
15:35
<
nats` >
in generate_dump_after
15:41
_whitelogger has joined #symbiflow
15:43
<
nats` >
yep so that but it failed on my test because I didn't have it :)
15:43
<
nats` >
fixed like your PR and rerun :)
16:22
<
nats` >
just output of the tcl script I made to fix (bad directory) and I should be good :)
16:22
<
nats` >
the final generation seems to work
16:22
<
nats` >
2019-01-15 17:18:51.938046 Generating reduced tile for CLBLM_L
16:22
<
nats` >
2019-01-15 17:18:51.938328 Using pool.imap_unordered
16:22
<
nats` >
100% (600 of 600) |###################################################################################################################################################################################| Elapsed Time: 0:02:00 Time: 0:02:00
16:22
<
nats` >
2019-01-15 17:21:10.662660 Generating reduced tile for CLBLM_R
16:22
<
nats` >
I think I'll try to do the same progress bar for the multithread version of fuzzer
16:23
<
nats` >
it's better than nothing or ton of line of text
16:35
powerbit has quit [Ping timeout: 272 seconds]
16:40
<
nats` >
the reducing stage is awfully long too !
16:40
<
nats` >
litghost, I think digshadow is right it's not possible to put 072 and 074 before other fuzzers... the 074 reduce stage is a nightmare
17:05
citypw has quit [Ping timeout: 250 seconds]
17:49
kraiskil has quit [Quit: Leaving]
17:57
<
nats` >
2019-01-15 18:07:30.598128 Creating wire<->node index
17:57
<
nats` >
100% (3824221077 of 3824221077) |#####################################################################################################################################################################| Elapsed Time: 0:00:59 Time: 0:00:59
17:57
<
nats` >
100% (5153171150 of 5153171150) |#####################################################################################################################################################################| Elapsed Time: 0:01:00 Time: 0:01:00
17:57
<
nats` >
2019-01-15 18:09:33.945337 Creating node tree
17:57
<
nats` >
35% (398115 of 1122477) |############################################################# | Elapsed Time: 0:00:07 ETA: 0:00:15Traceback (most recent call last):
17:57
<
nats` >
File "create_node_tree.py", line 279, in <module>
17:57
<
nats` >
File "create_node_tree.py", line 271, in main
17:57
<
nats` >
if node in uphill_wire_node_index else [])))
17:57
<
nats` >
File "create_node_tree.py", line 103, in create_ordered_wires_for_node
17:57
<
nats` >
assert len(wires_in_node) >= len(all_wires)
17:58
<
nats` >
AssertionError
17:58
<
nats` >
100% (1122477 of 1122477) |###########################################################################################################################################################################| Elapsed Time: 0:00:10 Time: 0:00:10
17:58
<
nats` >
something failed
18:33
<
nats` >
I guess I may need some help to sort that
18:33
<
nats` >
it seems to be unhappy of results of the 072 fuzzer and something else
18:38
<
litghost >
nats: Do put up the PR
18:38
<
litghost >
I'll take a look
19:35
<
tpb >
Title: WIP: reworking Fuzzer 074 for multi thread process - Dont MERGE ! by natsfr · Pull Request #526 · SymbiFlow/prjxray · GitHub (at github.com)
19:35
<
nats` >
not for merging until we are sure :)
19:36
<
nats` >
rubber ducky :D
19:36
<
nats` >
I think it fails because the 072 was generated for artix 7 and the 074 for zynq
19:39
<
litghost >
ya, that'll totally bork it
19:39
<
litghost >
FYI WIP: in the title is enough to prevent a merge, so you don't need a giant "don't merge", that's what WIP means
19:40
<
nats` >
ah oky sorry
19:40
<
nats` >
I wante dot be sur e:)
19:40
<
nats` >
I wanted to be sure
19:41
<
nats` >
I rerun a full 072/074
19:41
<
nats` >
but feel free to test :)
19:41
<
nats` >
do you remember how long is the 074 run ?
19:50
<
litghost >
I usually leave it overnight
19:51
<
litghost >
There is nothing in the processing that truely requires that length, it is just severly unoptimized
19:52
<
litghost >
Let's first fix the vivado leak per your PR, then maybe consider attacking the post-processing script. I believe it should be possible to speed it up significantly with only small modifications
20:21
<
nats` >
I didn't look into that
20:22
<
nats` >
but I was wondering if the json processing is optimized
20:22
<
nats` >
I have pretty bad souvenir of json lib
20:22
<
nats` >
but that was really long time ago
20:22
<
nats` >
anyway all the vivado processing for zynq take less than 50 minute for the 074
20:23
<
nats` >
and I started one with time to see the full process
20:24
<
litghost >
"but I was wondering if the json processing is optimized" -> Depends on what you mean. The json parsing library is reasonable fast, but problem I believe is some inefficient data passing methods. When I wrote the post-processing for 074, I proritized something working, so it isn't as clean as it might have been. Some profiling and refactoring would probably go a long way.
20:26
<
nats` >
I guess you'll be more efficient than me as you could see my python knowledge is long outdated now
20:26
<
nats` >
so performance wise..... :D
20:26
<
litghost >
reducing the runtime and memory usage of the vivado portion is of immediate benefit, so the PR is still very valuable
20:27
<
nats` >
sure sure :)
20:27
<
nats` >
and I'm happy to help on that
20:51
<
nats` >
uhhmmmmm got an idea !
20:51
<
nats` >
maybe I can make a bug report on that
20:51
<
nats` >
we should store in the run.ok the target of the last run !
20:51
<
nats` >
you would avoid stupid mistake like I made
20:54
<
tpb >
Title: Fuzzer storing last target · Issue #528 · SymbiFlow/prjxray · GitHub (at github.com)
21:08
<
nats` >
uhhhh litghost the post processing is about to blow my RAM away :D
21:08
<
nats` >
it seems it's really tight in 16GB
21:13
<
litghost >
nats: Ya, that seems about right
21:14
<
litghost >
I'm currently working on testing a couple INT pip CLs, and I'll move over to checking out your PR
21:14
<
nats` >
I'll try to get 16GB more next month
21:14
<
nats` >
it's too expensive for this one :D
21:26
<
kgugala >
nats`: I also started a 072/074 build
21:28
<
nats` >
with the PR ?
21:28
<
nats` >
and which target ?
21:28
<
kgugala >
should I try zynq instead?
21:28
<
nats` >
should work on all
21:28
<
nats` >
the artix is longer
21:29
<
nats` >
zynq < artix < kintex
21:29
<
nats` >
count about 30minute 072 for artix
21:29
<
nats` >
and I would say 1h15 074 BEFORE post processing
21:29
<
nats` >
I couldn't run the post proc
21:30
<
nats` >
I'm trying it for zynq right now
21:31
<
nats` >
I think we really have to stabilize all of that and generate as few as possible
21:31
<
nats` >
uhhmm how hard would it be to distribute the build over network
21:31
<
nats` >
there are python lib for that ?
21:43
<
nats` >
duhhh we really need to optimise the post processing
21:44
<
nats` >
it's awfully slow :D
22:05
<
litghost >
In general 072-074 weren't optimized because they are very deterministic, and don't need to be re-run often.
22:05
<
litghost >
This is unlike the other fuzzers which are fairly stochastic
22:08
<
nats` >
Traceback (most recent call last):
22:08
<
nats` >
File "generate_grid.py", line 676, in <module>
22:08
<
nats` >
File "generate_grid.py", line 578, in main
22:08
<
nats` >
with open(os.path.join(prjxray.util.get_db_root(), 'tilegrid.json')) as f:
22:08
<
nats` >
FileNotFoundError: [Errno 2] No such file or directory: '/home/nats/project/symbiflow/prjxray/database/zynq7/tilegrid.json'
22:08
<
nats` >
Makefile:16: recipe for target 'build/specimen_001/OK' failed
22:08
<
nats` >
make[1]:
*** [build/specimen_001/OK] Error 1
22:08
<
nats` >
make[1]: Leaving directory '/home/nats/project/symbiflow/prjxray/fuzzers/074-dump_all'
22:08
<
nats` >
Makefile:20: recipe for target 'run' failed
22:08
<
nats` >
make:
*** [run] Error 2
22:08
<
nats` >
Command exited with non-zero status 2
22:08
<
nats` >
Command being timed: "make run"
22:08
<
nats` >
User time (seconds): 17837.88
22:09
<
nats` >
System time (seconds): 1258.89
22:09
<
nats` >
Percent of CPU this job got: 305%
22:09
<
nats` >
Elapsed (wall clock) time (h:mm:ss or m:ss): 1:44:09
22:09
<
nats` >
Average shared text size (kbytes): 0
22:09
<
nats` >
Average unshared data size (kbytes): 0
22:09
<
nats` >
Average stack size (kbytes): 0
22:09
<
nats` >
Average total size (kbytes): 0
22:09
<
nats` >
Maximum resident set size (kbytes): 4110148
22:09
<
nats` >
Average resident set size (kbytes): 0
22:09
<
nats` >
Major (requiring I/O) page faults: 44
22:09
<
nats` >
Minor (reclaiming a frame) page faults: 659168440
22:09
<
nats` >
Voluntary context switches: 5940756
22:09
<
nats` >
Involuntary context switches: 10981805
22:09
<
nats` >
File system inputs: 13773992
22:09
<
nats` >
File system outputs: 18203112
22:09
<
nats` >
Socket messages sent: 0
22:09
<
nats` >
Socket messages received: 0
22:09
<
nats` >
Signals delivered: 0
22:09
<
nats` >
Page size (bytes): 4096
22:09
<
nats` >
Exit status: 2
22:09
<
nats` >
ar sorry bad copy paste
22:09
<
nats` >
it failed at 1h44 -_-
22:09
<
nats` >
I think we really should do a sanity check before running
22:11
<
nats` >
ahhh I didn't run previous fuzzer
22:12
<
tpb >
Title: Fuzzer pre-run sanity check · Issue #529 · SymbiFlow/prjxray · GitHub (at github.com)
22:15
<
nats` >
oky I started a full run of all fuzzer from the main directory
22:15
<
nats` >
it'll run all night hopefully
22:15
<
nats` >
good night !
22:17
<
nats` >
uhhmm it stopped right before going to be
22:17
<
nats` >
I'll check that tmorrow
23:53
tpb has quit [Remote host closed the connection]
23:53
tpb has joined #symbiflow