sb0 changed the topic of #m-labs to: https://m-labs.hk :: Mattermost https://chat.m-labs.hk :: Logs http://irclog.whitequark.org/m-labs
spyder has quit [Ping timeout: 256 seconds]
Gurty has quit [Excess Flood]
Gurty has joined #m-labs
gnufan_home has quit [Ping timeout: 258 seconds]
gnufan_home has joined #m-labs
gnufan_home has quit [Ping timeout: 245 seconds]
gnufan_home has joined #m-labs
gnufan_home has quit [Ping timeout: 268 seconds]
gnufan_home has joined #m-labs
gnufan_home has quit [Ping timeout: 248 seconds]
proteusguy has quit [Ping timeout: 244 seconds]
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
<_whitenotifier-3> [nmigen] whitequark opened issue #86: Platform clock definitions are very inflexible - https://git.io/fjuVm
<_whitenotifier-3> [nmigen] whitequark commented on issue #86: Platform clock definitions are very inflexible - https://git.io/fjuV3
<_whitenotifier-3> [nmigen] cr1901 commented on issue #86: Platform clock definitions are very inflexible - https://git.io/fjuVc
<_whitenotifier-3> [nmigen] whitequark commented on issue #86: Platform clock definitions are very inflexible - https://git.io/fjuVW
<_whitenotifier-3> [nmigen] whitequark opened issue #87: Automatic definition of false path constraints - https://git.io/fjuVR
<_whitenotifier-3> [m-labs/nmigen] whitequark pushed 2 commits to master [+0/-0/±3] https://git.io/fjuVz
<_whitenotifier-3> [m-labs/nmigen] whitequark 452c4b3 - vendor.lattice_ice40: normalize device names.
<_whitenotifier-3> [m-labs/nmigen] whitequark c52cd72 - Typos and style fixes. NFC.
gnufan_home has joined #m-labs
<_whitenotifier-3> [nmigen] Success. The Travis CI build passed - https://travis-ci.org/m-labs/nmigen/builds/541536215?utm_source=github_status&utm_medium=notification
<_whitenotifier-3> [nmigen] Success. 80.71% remains the same compared to 4379a5d - https://codecov.io/gh/m-labs/nmigen/commit/c52cd72d3e19716102ed7815923d6c4427c205ec
<_whitenotifier-3> [nmigen] Success. Coverage not affected when comparing 4379a5d...c52cd72 - https://codecov.io/gh/m-labs/nmigen/commit/c52cd72d3e19716102ed7815923d6c4427c205ec
gnufan_home has quit [Ping timeout: 272 seconds]
proteusguy has joined #m-labs
gnufan_home has joined #m-labs
gnufan_home has quit [Ping timeout: 248 seconds]
gnufan_home has joined #m-labs
proteusguy has quit [Ping timeout: 272 seconds]
rohitksingh has joined #m-labs
proteusguy has joined #m-labs
m4ssi has joined #m-labs
proteusguy has quit [Ping timeout: 248 seconds]
<_whitenotifier-3> [m-labs/nmigen] whitequark pushed 2 commits to master [+0/-0/±14] https://git.io/fjuKu
<_whitenotifier-3> [m-labs/nmigen] whitequark ab3f103 - build.dsl: replace extras= with Attrs().
<_whitenotifier-3> [m-labs/nmigen] whitequark c9879c7 - build.{dsl,res,plat}: apply clock constraints to signals, not resources.
<_whitenotifier-3> [m-labs/nmigen-boards] whitequark pushed 2 commits to master [+0/-0/±7] https://git.io/fjuKz
<_whitenotifier-3> [m-labs/nmigen-boards] whitequark e8dcc1e - Use vendor's device and package names everywhere, not nextpnr's.
<_whitenotifier-3> [m-labs/nmigen-boards] whitequark ec6316e - Update to track changes in nmigen.
<_whitenotifier-3> [nmigen] whitequark closed issue #86: Platform clock definitions are very inflexible - https://git.io/fjuVm
<_whitenotifier-3> [nmigen] Success. The Travis CI build passed - https://travis-ci.org/m-labs/nmigen/builds/541631711?utm_source=github_status&utm_medium=notification
<_whitenotifier-3> [nmigen] Success. Absolute coverage decreased by -0.19% but relative coverage increased by +2.33% compared to c52cd72 - https://codecov.io/gh/m-labs/nmigen/commit/c9879c795bf7d5ad2e34106c0cf1adc6a22e070b
<_whitenotifier-3> [nmigen] Success. 83.05% of diff hit (target 80.71%) - https://codecov.io/gh/m-labs/nmigen/commit/c9879c795bf7d5ad2e34106c0cf1adc6a22e070b
futarisIRCcloud has joined #m-labs
rohitksingh has quit [Ping timeout: 268 seconds]
sb0 has joined #m-labs
<sb0> whitequark: do you know a nice way to convert vectorworks to parasolid x_t, or is it TPB + windows VM?
<whitequark> i've never even touched vectorworks or parasolid
<hartytp> whitequark: looking at bt/disasm, a lot of code seems to be inlined into main. do you have any pointers for how to fix that/get a better diagnostic
sb0 has quit [Quit: Leaving]
<whitequark> hartytp: gdb should be able to show inlined functions in the backtrace
<whitequark> are you generating/using debug information?
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
<hartytp> whitequark: yes https://github.com/quartiq/stabilizer/blob/a575611e82edf236abba4cb6c6f8eec30caeadbb/Cargo.toml#L61 (playing around with rjo's stabilizer code and learning-through-breaking)
<whitequark> hartytp: hm, well you could reduce opt-level or sprinkle #[noinline]
<whitequark> but i really do wonder why gdb isn't showing you inlined functions in the backtrace
<hartytp> I seem to get better results without lto
<whitequark> well yes, lto opens more opportunities for inlining
<whitequark> same effect as reducing opt-level
<zignig> whitequark: just looking over the changes to the clocks on the new platform system.
<zignig> what was the reasoning behind removing the plaform.clock member ?
<zignig> there does not seem to be a way to get a list of the available clocks on a platform without having them assinged already.
<hartytp> whitequark: fwiw, here is the kind of bt I'm seeing https://paste.debian.net/1086364/ not sure where to begin with this, anyway, thanks for the help
<whitequark> zignig: you could not define a clock constraint on something like "eth#0.rx_clk"
<whitequark> i.e. a subsignal of some resource
<whitequark> not even easily define, define at all
<whitequark> it's possible to add some ad-hoc language for specifying subresources, but i decided that instead clocks should be defined by association with their Pin or (if you're not requesting a Pin) raw inout wire
<whitequark> that you could enumerate (some) clocks was more of an artifact of the system i changed than its purpose
<zignig> ok I see where you are coming from. So far all my designs are simplistic ( still getting the hang of RTL )
<whitequark> oMigen had something called default_clk_name and default_clk_period
<whitequark> I don't like it a whole lot either...
<whitequark> I might be able to allow requesting clock constraints (if any) from the platform, that would solve half of the problem
<zignig> Thanks for all you work on nmigen, I'm having fun hacking on my tinybx.
<zignig> I have a uart ported from your orignal blog post and I am attempting to hook it up to a Boneless-CPU
<whitequark> that UART isn't especially good btw, it's one of my earliest attempts
<whitequark> today i would implement it in a completely different (and much simpler) way
<zignig> it works at 57600 :)
<whitequark> mostly just a shift register
<whitequark> and no oversampling
<zignig> is there going to be a nmigen-cores at some point ?
<whitequark> probably
<zignig> :)
<zignig> I keep on getting "Warning: No clocks found in designWarning: No clocks found in design2 warnings, 0 errors" when I add a boneless CPU to a platformed script
<whitequark> is your nextpnr up to date?
<zignig> yep, I've go nextpnr and yosys updating on a cron job...
<whitequark> archive the build directory and upload it somewhere then
<zignig> ok
<_whitenotifier-3> [m-labs/nmigen] whitequark pushed 1 commit to master [+0/-0/±2] https://git.io/fjuXr
<_whitenotifier-3> [m-labs/nmigen] whitequark b45c511 - build.res: allow querying frequency of a previously constrained clock.
<_whitenotifier-3> [nmigen] Success. The Travis CI build passed - https://travis-ci.org/m-labs/nmigen/builds/541724681?utm_source=github_status&utm_medium=notification
<_whitenotifier-3> [nmigen] Success. 80.6% (+0.08%) compared to c9879c7 - https://codecov.io/gh/m-labs/nmigen/commit/b45c5119f573e156a3aad4a362ca210d6e40505e
<_whitenotifier-3> [nmigen] Success. 100% of diff hit (target 80.52%) - https://codecov.io/gh/m-labs/nmigen/commit/b45c5119f573e156a3aad4a362ca210d6e40505e
<zignig> whitequark: build folder , https://bl3dr.com/boneless_clock_issue.tgz
<zignig> whitequark: you might like my other project too. https://cqparts.github.com/examples/ , 3d stuff in python.
cr1901_modern has quit [Quit: Leaving.]
<plaes> zignig: your last link is 404
<plaes> whitequark: regarding m-labs/openocd repository in github, I rebased it against upstream repository and ended up with only two extra commits on-top of master: https://github.com/plaes/openocd/commits/m-labs-master
<whitequark> plaes: i don't work on that repository
<whitequark> cc rjo
<zignig> erk , whoops.
* zignig reflexy typing fingers added .com , grr
<_whitenotifier-3> [m-labs/nmigen-boards] whitequark pushed 1 commit to master [+0/-0/±4] https://git.io/fjuMx
<_whitenotifier-3> [m-labs/nmigen-boards] whitequark c6e28e8 - Simplify after changes in nmigen.
<whitequark> sb: ping
<_whitenotifier-3> [nmigen] whitequark opened issue #88: Signal names are not translated for constraints, etc. - https://git.io/fjuDl
rohitksingh has joined #m-labs
sb0 has joined #m-labs
<sb0> whitequark: how do you build rust-lld?
<whitequark> sb0: what exactly do you mean by rust-lld?
<sb0> whatever it's looking for when it craps out with
<sb0> error: linker `rust-lld` not found
<sb0> afaict it's something that is compiled when rustc is compiled
<whitequark> how do you build rustc?
<whitequark> how are you building*
<sb0> it's the nixos rustc and llvm, except I am passing "-DLLVM_EXPERIMENTAL_TARGETS_TO_BUILD=RISCV" to llvm
<sb0> when compiling it
<sb0> I want to build a bare metal riscv firmware. should I be using this rustc-lld? I can also work around the problem with -C linker=xxx
<whitequark> let me look
<sb0> but by default (and with a minimal cargo configuration) it attempts to use this non-existent rust-lld
<whitequark> sb0: configure --enable-lld, i think
<sb0> that's it? why is it not the default when it attempts to use it by default?
<whitequark> you're supposed to use --enable-full-tools, i think
<sb0> hmm that seems to do ...something. waiting for rustc to be compiled
rohitksingh has quit [Ping timeout: 272 seconds]
proteusguy has joined #m-labs
rohitksingh has joined #m-labs
hartytp has quit [Quit: Page closed]
cr1901 has joined #m-labs
rohitksingh has quit [Ping timeout: 272 seconds]
lkcl has quit [Read error: Connection reset by peer]
lkcl has joined #m-labs
<cr1901> whitequark: When trying to run the blinky in this platform file I' m making (by attaching a PMOD): http://ix.io/1L1O, I get the following error: http://ix.io/1L1T
<cr1901> Would you have any idea what's causing this error?
<cr1901> I would think the pmod connector exists since I defined it in "connectors"
stekern has quit [Ping timeout: 245 seconds]
lynxis has quit [Remote host closed the connection]
lynxis has joined #m-labs
stekern has joined #m-labs
mumptai has joined #m-labs
<rjo> plaes: ack. that looks correct.
mumptai has quit [Remote host closed the connection]
X-Scale` has joined #m-labs
<zignig> cr1901: having a look at your code... there is no pin 5 on pmod 0. user_led 0 is bound to nothing.
X-Scale has quit [Ping timeout: 272 seconds]
[X-Scale] has joined #m-labs
[X-Scale] is now known as X-Scale
X-Scale` has quit [Ping timeout: 245 seconds]
<cr1901> Oh you can't be serious... you actually have to map to the actual pin numbers on the connector?
<cr1901> as opposed to 0, 1, 2, 3, 4, 5, etc...
<cr1901> zignig: Yea I don't get this... how do I map user_led 0 to the fifth pin on the pmod2 connector?
<zignig> hmmm , sorry my bad, it should map the ordinal of the pin from the connector.
<zignig> however I'm not sure what it does with pins marked '-'
<cr1901> zignig: They are counted
<zignig> your platform , from platform.connectors in the console
<cr1901> which means that pin "5" should become "7"
<zignig> (('pmod', 0), (connector pmod 0 1=>4 2=>2 3=>47 4=>45 7=>3 8=>48 9=>46 10=>44))
<cr1901> and "6" becomes "8"
<zignig> so the 5th ping is 7
<cr1901> yea... and this makes sense in retrospect, but still throws me off
<zignig> indeed ;)
<zignig> bashing about and finding other ways to do things is a important part of nMigen...
<cr1901> And with that problem solved, I'm ready to make a PR for icebreaker
<_whitenotifier-3> [nmigen-boards] cr1901 opened pull request #2: Add icebreaker platform. - https://git.io/fjubz
<zignig> nice, more boards.
<zignig> I'm trying to get a Boneless together for a platformed tinybx at the moment.
<zignig> slow going , but getting closer.