00:19
antgreen` has joined #m-labs
00:19
antgreen has quit [Remote host closed the connection]
00:20
ylamarre has quit [Ping timeout: 250 seconds]
04:18
ylamarre has joined #m-labs
05:30
ylamarre has quit [Ping timeout: 240 seconds]
06:36
acathla has quit [Quit: Coyote finally caught me]
06:54
acathla has joined #m-labs
07:55
<
sb0 >
how to get git to re-read .gitmodules?
07:55
<
sb0 >
and git submodule init doesn't help
07:55
<
whitequark >
git submodule update
07:57
<
whitequark >
bizarre
07:59
<
sb0 >
I guess the usual solution to git problems (rm -rf + clone) applies ...
08:07
<
sb0 >
hmm. apparently they needed another "git submodule add", because they had been "git rm"'d while I moved things around
08:07
<
sb0 >
interestingly, the git rm didn't update .gitmodules or .git/config
08:44
rohitksingh has joined #m-labs
08:47
<
sb0 >
yay, new-misoc boots on ppro
08:49
<
GitHub183 >
misoc/new da425d1 Sebastien Bourdeauducq: add stream, fix CPUs and more imports. simple target boots on ppro.
08:49
<
GitHub112 >
migen/new 6790349 Sebastien Bourdeauducq: fhdl: export DUID
08:49
<
GitHub112 >
migen/new 8f42b6f Sebastien Bourdeauducq: fhdl: introduce wrap function
08:49
<
GitHub112 >
migen/new fa1e8cd Sebastien Bourdeauducq: wrap expressions in Specials
10:13
rohitksingh has quit [Ping timeout: 264 seconds]
10:54
<
GitHub105 >
misoc/new 67133f3 Sebastien Bourdeauducq: replace flen with len
10:54
<
GitHub97 >
migen/new 808cf06 Sebastien Bourdeauducq: fhdl: replace flen with len
11:26
<
mithro >
hey sb0 - is migen support to be deterministic in the verilog it generates?
11:50
<
mithro >
Looks like something is looping over a dictionary because if I set PYTHONHASHSEED then the output becomes the same
12:12
<
mithro >
hrm, it seems to be something the HDMI module is doing....
12:13
antgreen` has quit [Ping timeout: 246 seconds]
12:14
<
mithro >
sb0: any suggestions on how to track it down?
12:14
<
sb0 >
...look at how the unstable code is generated? ;)
12:14
<
sb0 >
what Module it comes from, how it works, etc.
12:19
<
mithro >
actually - it seems to be something in the sdram...
12:36
antgreen` has joined #m-labs
13:33
ylamarre has joined #m-labs
13:45
<
mithro >
sb0: any ideas? it seems the examples/basic/memory.py exhibits this problem...
13:46
<
mithro >
sb0: same as psync.py
13:48
<
sb0 >
mithro, not off the top of my head.
13:49
<
sb0 >
which branch is that affecting?
13:49
<
mithro >
sb0: master
13:50
<
mithro >
migen/examples/basic$ rm -f *.v; for p in memory.py psync.py; do for i in 1 2; do PYTHONHASHSEED=$i python $p > $p.$i.v; done; done; md5sum *.v
13:51
<
mithro >
the Memory example uses a dictionary for the self.ios
13:51
<
sb0 >
you mean a set?
13:51
<
mithro >
sorry, yes
13:52
<
sb0 >
well, it should handle this case deterministically. if not, consider sending a patch.
13:52
<
sb0 >
(that's what the .duid attributer is for)
13:53
<
GitHub198 >
migen/new e136352 Sebastien Bourdeauducq: fhdl/structure: relax type requirements for Array elements
13:53
<
GitHub198 >
migen/new 09003a5 Sebastien Bourdeauducq: fhdl/specials/Tristate: handle i=None
13:54
<
sb0 >
mithro, btw, your 'shutil.rmtree("build")' messes with git
13:54
<
sb0 >
there's build/.keep_me ...
13:54
<
mithro >
sb0: yeah, I discovered that recently too
13:56
<
sb0 >
tried reverting, git tells me there are conflicts, but reverted it ok without conflict markers
13:57
<
sb0 >
oh, no, it's more aftermath of the submodule fuckup
13:58
<
GitHub181 >
misoc/new 27b2383 Sebastien Bourdeauducq: sdram working on PPro
13:58
<
GitHub181 >
misoc/new a186bfe Sebastien Bourdeauducq: Revert "Use shutil rather then rm -rf command."...
13:59
<
GitHub110 >
misoc/master 4fe0f60 Sebastien Bourdeauducq: Revert "Use shutil rather then rm -rf command."...
14:07
<
mithro >
sb0: well for the examples, I can fix the issue by the following patch -
14:10
<
mithro >
sb0: but I don't think that is the right patch
14:14
<
mithro >
rm -rf 1 2; mkdir 1; mkdir 2; for p in $(grep -l "verilog.convert(" *.py); do for i in 1 2; do export PYTHONHASHSEED=$i; python $p > $i/$p.v; done; done; diff -s -u -r 1 2
14:14
<
mithro >
that is what I'm using to test now
14:32
travis-ci has joined #m-labs
14:32
<
travis-ci >
m-labs/migen#110 (new - fa1e8cd : Sebastien Bourdeauducq): The build was broken.
14:32
travis-ci has left #m-labs [#m-labs]
14:32
travis-ci has joined #m-labs
14:32
<
travis-ci >
m-labs/migen#112 (new - 09003a5 : Sebastien Bourdeauducq): The build passed.
14:32
travis-ci has left #m-labs [#m-labs]
14:41
rohitksingh has joined #m-labs
15:17
<
mithro >
sb0: sorry, I guess that should be on the migen repo, not the misoc repo?
16:28
mumptai has joined #m-labs
17:35
rohitksingh has quit [Ping timeout: 250 seconds]
17:49
rohitksingh has joined #m-labs
17:49
antgreen` has quit [Read error: Connection reset by peer]
18:46
antgreen has joined #m-labs
21:06
stekern has quit [Ping timeout: 264 seconds]
21:08
stekern has joined #m-labs
21:43
rohitksingh has quit [Quit: Leaving.]
22:58
mumptai has quit [Quit: Verlassend]
23:54
stekern has quit [Ping timeout: 240 seconds]
23:56
stekern has joined #m-labs