Skip to content

Commit 4f88873

Browse files
bors[bot]japaric
bors[bot]
and
japaric
committed
Merge #240
240: add minutes and IRC logs of the last meeting r=jamesmunns a=japaric Co-authored-by: Jorge Aparicio <[email protected]>
2 parents 6da5b97 + 18977f0 commit 4f88873

File tree

2 files changed

+536
-0
lines changed

2 files changed

+536
-0
lines changed

‎minutes/2018-10-09.irc.log

Lines changed: 165 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,165 @@
1+
2018-10-09 20:00:15 ~japaric meeting agenda: https://paper.dropbox.com/doc/Embedded-WG--AO3FUbH62OMOdbN4jsAV9KvQAg-5pdv734N8KpxHFMJijuoL
2+
2018-10-09 20:00:22 @adamgreig really i want cargo features with arguments :P
3+
2018-10-09 20:00:26 jamesmunns the linker can't change the size of an array, afaik
4+
2018-10-09 20:00:36 jamesmunns adamgreig: yeah, that would be mostly better
5+
2018-10-09 20:00:48 @adamgreig assuming you can somehow use the argument as a const
6+
2018-10-09 20:00:49 jamesmunns dylan-dpc was talking about submitting an RFC to address that
7+
2018-10-09 20:01:04 @adamgreig dread to think how it would interoperate with the current additive-only model though
8+
2018-10-09 20:01:14 @adamgreig what do you do when two crates depend on the same third crate but with different arguments?
9+
2018-10-09 20:01:15 jamesmunns adamgreig: I was hoping it was possible to use a const in a cfg! block, but I don't think thats a thing at all
10+
2018-10-09 20:01:17 @adamgreig nightmare
11+
2018-10-09 20:01:36 jamesmunns adamgreig: yeah, I think thats why they went with the union of all features
12+
2018-10-09 20:01:39 hannobraun Good evening
13+
2018-10-09 20:01:48 jamesmunns you might need a separate interface than current cfg
14+
2018-10-09 20:01:54 @adamgreig yea. still i'd love to have it.
15+
2018-10-09 20:04:32 ~japaric ok, let's start this meeting
16+
2018-10-09 20:04:50 ~japaric we are 8 weeks away from the edition release; let's check on the edition issues
17+
2018-10-09 20:05:18 ~japaric issue 1: book intro; anything to report?
18+
2018-10-09 20:05:32 @theJPster sorry, is there a link? I'm on a different PC today.
19+
2018-10-09 20:05:45 jamesmunns https://paper.dropbox.com/doc/Embedded-WG--AO30Ou0UvgAU8jWWKY7CN6QLAg-5pdv734N8KpxHFMJijuoL
20+
2018-10-09 20:06:34 ~japaric looks like there's a PR to update the intro https://github.com/rust-embedded/book/pull/40
21+
2018-10-09 20:07:01 ~japaric and looks like it was updated and it's waiting for re-review
22+
2018-10-09 20:07:29 jamesmunns I can look at it tomorrow, if no-one has time before then
23+
2018-10-09 20:07:33 @therealprof Looks good to me on first glance.
24+
2018-10-09 20:07:38 @therealprof I'll check later.
25+
2018-10-09 20:08:09 ~japaric moving on, issue 2: the first step chapter
26+
2018-10-09 20:08:18 ~japaric steps*
27+
2018-10-09 20:08:28 ~japaric no news on this from my side
28+
2018-10-09 20:09:44 @theJPster I find myself with some spare time this week
29+
2018-10-09 20:10:03 @theJPster I need to get ready for RBR but if there's something it would be useful to spend, say, six hours on, just point me at it.
30+
2018-10-09 20:10:29 ~japaric chapter 2 is missing content on semihosting and register access
31+
2018-10-09 20:10:58 @theJPster I can do that.
32+
2018-10-09 20:11:03 ~japaric theJPster: +1
33+
2018-10-09 20:12:07 ~japaric issue 3, collections chapter: no change since last week
34+
2018-10-09 20:12:30 ~japaric section 4, concurrency. any news adamgreig?
35+
2018-10-09 20:12:36 @adamgreig hope to have a pr this evening
36+
2018-10-09 20:12:52 @adamgreig hoped to have a pr for the meeting but not quite :p
37+
2018-10-09 20:13:10 ~japaric adamgreig: nice
38+
2018-10-09 20:13:57 ~japaric issues 5 and 6, portability and type state
39+
2018-10-09 20:14:03 ~japaric I think I saw some PRs
40+
2018-10-09 20:14:06 jamesmunns PRs are open
41+
2018-10-09 20:14:21 jamesmunns type state is on my plan this week, portability is likely not
42+
2018-10-09 20:14:31 jamesmunns I think someone volunteered for portability last week though
43+
2018-10-09 20:14:56 * japaric tries to remember
44+
2018-10-09 20:16:41 jamesmunns I'll skim through the logs in a bit
45+
2018-10-09 20:17:30 ~japaric I looked at the logs and didn't spot any volunteer for portability stuff
46+
2018-10-09 20:17:50 ~japaric issue 7, tips for C/C++. adamgreig?
47+
2018-10-09 20:17:53 @adamgreig korken89 and i were going to work on tips but they're away this week; i plan to put some bits together this week after concurrency
48+
2018-10-09 20:18:14 @adamgreig nothing actually written yet afaik
49+
2018-10-09 20:18:39 ~japaric ok, thanks for the update
50+
2018-10-09 20:19:28 ~japaric issue 8, C/Rust integration
51+
2018-10-09 20:19:38 @adamgreig pr still waiting on author
52+
2018-10-09 20:19:56 @adamgreig spacekookie: ^?
53+
2018-10-09 20:20:19 jamesmunns I was hoping spacekookie would have had time this past week, but hasn't. I will follow up with her to see if I can close it out
54+
2018-10-09 20:20:30 jamesmunns Feel free to assign to me
55+
2018-10-09 20:20:49 jamesmunns (at least closing out that PR)
56+
2018-10-09 20:20:54 ~japaric jamesmunns: +1
57+
2018-10-09 20:21:53 ~japaric issue 9, outtline: no change
58+
2018-10-09 20:22:34 ~japaric issue 10, the other book
59+
2018-10-09 20:22:39 ~japaric I saw some PRs
60+
2018-10-09 20:23:27 ~japaric guess they are already merged
61+
2018-10-09 20:24:07 ~japaric I think ryankurte and korken89 are not around
62+
2018-10-09 20:25:25 ~japaric issue 11, the showcase
63+
2018-10-09 20:25:48 ~japaric I was hoping to take a stab at it during last weekend but ended having no time
64+
2018-10-09 20:26:04 ~japaric won't have time this week either ...
65+
2018-10-09 20:27:04 ~japaric issue 12, tests in rust-lang/rust
66+
2018-10-09 20:27:33 ~japaric there's an open PR but turns out we can't use the proc-macro crate in run-make tests for targets (that are not the host)
67+
2018-10-09 20:28:00 ~japaric so the PR will have to be updated to use cortex-m-rt v0.5.4 which uses the old macro_rules! macros (entry!)
68+
2018-10-09 20:28:27 @therealprof :(
69+
2018-10-09 20:28:54 @therealprof No fix in sight?
70+
2018-10-09 20:29:16 @theJPster :(
71+
2018-10-09 20:29:37 ~japaric for the proc-macro stuff? there's a PR that should fix the issue and have been open for several weeks but it's complicated to land
72+
2018-10-09 20:30:34 ~japaric PR: https://github.com/rust-lang/rust/pull/49219
73+
2018-10-09 20:30:57 ~japaric s/several weeks/several months/
74+
2018-10-09 20:31:54 ~japaric using the older version of cortex-m-rt is the only way to do the job in time for 1.31
75+
2018-10-09 20:32:24 @theJPster are there any downsides, apart from being a bit ugly?
76+
2018-10-09 20:32:46 ~japaric theJPster: no, the linker stuff is the same; only the UI is different
77+
2018-10-09 20:33:14 ~japaric in terms of what will be tested there's no different between that version and v0.6.x
78+
2018-10-09 20:33:34 @theJPster ok
79+
2018-10-09 20:33:49 @theJPster something is better than nothing
80+
2018-10-09 20:34:02 jamesmunns Still covers the things we want to test
81+
2018-10-09 20:34:24 ~japaric next: issue 19, docs.rust-embedded.org
82+
2018-10-09 20:34:24 jamesmunns As proc macro generation is probably not likely to specifically break embedded
83+
2018-10-09 20:34:51 ~japaric this is not blocked on decision making and can already happen
84+
2018-10-09 20:35:03 ~japaric I'll ping nastevens
85+
2018-10-09 20:36:29 ~japaric re: rename; given that there's majority we should go ahead with it. But we should it at the same time docs.r-e.o goes live to minimize breakage
86+
2018-10-09 20:37:41 jamesmunns :+1:
87+
2018-10-09 20:38:16 ~japaric last: issue 20, webpage. I think we are on track
88+
2018-10-09 20:38:37 jamesmunns oh. right. I think there were some review questions for me to close out
89+
2018-10-09 20:38:45 jamesmunns should be quick, will add to this week's todo
90+
2018-10-09 20:38:53 ~japaric +1
91+
2018-10-09 20:39:22 ~japaric jamesmunns: perhaps you can update the screenshot on the internal team chat (after the PR is merged) so others can see the final state
92+
2018-10-09 20:40:15 jamesmunns I need to un-set my system fonts, it makes the website look overly nice and monospace :)
93+
2018-10-09 20:40:29 jamesmunns but will do, adding a note
94+
2018-10-09 20:40:58 ~japaric ok, that covers the triage
95+
2018-10-09 20:41:12 ~japaric this last part of the meeting can be used to share updates
96+
2018-10-09 20:41:37 ~japaric as of today we have a new team: the cortex-a team :tada:
97+
2018-10-09 20:41:52 @adamgreig request for feedback or votes on the voting majority pr, https://github.com/rust-embedded/wg/pull/206
98+
2018-10-09 20:42:03 @adamgreig and japaric i think we can go ahead and merge https://github.com/rust-embedded/cortex-m-rt/pull/122 ?
99+
2018-10-09 20:42:33 @therealprof min_const_fn: PR for bare-metal now or as soon as next stable is out?
100+
2018-10-09 20:43:02 @therealprof adamgreig: Yes, please.
101+
2018-10-09 20:43:33 ~japaric adamgreig: feel free to push that towards the finish line; for me it's lower priority than working on the book (which I'll resume once I'm done with my non-Rust tasks)
102+
2018-10-09 20:43:34 @therealprof Needs conflict resolving, though.
103+
2018-10-09 20:44:10 ~japaric therealprof: technically, using min_const_fn is a breaking change because it bumps the minimum required version
104+
2018-10-09 20:44:51 @theJPster PR #122 - is there a matching change to svd2rust, or am I supposed to implement the interrupt checks in a <chip>-hal crate? It feels like something that should be auto-generated.
105+
2018-10-09 20:45:07 ~japaric so a new minor release on both bare-metal and cortex-m will be needed
106+
2018-10-09 20:45:16 @adamgreig theJPster: https://github.com/rust-embedded/svd2rust/pull/235
107+
2018-10-09 20:45:25 ~japaric theJPster: "The companion svd2rust PR is rust-embedded/svd2rust#235"
108+
2018-10-09 20:45:33 @theJPster lol
109+
2018-10-09 20:45:37 @therealprof japaric: Yes, I agree. And it would be d-n-m for now. Just a question of the right timing.
110+
2018-10-09 20:45:45 @theJPster In my defence I'm on a 720p TV right now...
111+
2018-10-09 20:46:23 jamesmunns It's better than a doubled 640x480 monotron display :D
112+
2018-10-09 20:46:39 jamesmunns (or is that 800x600?)
113+
2018-10-09 20:47:06 @theJPster It's 400x600. And actually, that might be more readable.
114+
2018-10-09 20:48:11 @therealprof For me that is pretty much THE blocker at the moment so I'm very psyched to have that in ASAP.
115+
2018-10-09 20:48:46 @therealprof I'm not sure it actually is a breaking change in bare-metal.
116+
2018-10-09 20:49:15 @therealprof I'd keep the feature for now and just make it a no-op.
117+
2018-10-09 20:49:15 ~japaric you can make it non-breaking if you put it behing a feature gate
118+
2018-10-09 20:50:16 @theJPster Sorry, what is min_const_fn?
119+
2018-10-09 20:50:40 ~japaric minimal subset of 'const fn' that's available on stable
120+
2018-10-09 20:50:40 @therealprof I believe the only breakage would be to remove the feature gate.
121+
2018-10-09 20:51:04 ~japaric it's actually what was specced in the original const-fn RFC
122+
2018-10-09 20:51:23 @theJPster oh nice. Monotron is stuck on nightly because of that.
123+
2018-10-09 20:51:26 ~japaric unstable const-fn includes stuff like panics, if-else, etc
124+
2018-10-09 20:51:56 @therealprof Which allows to define static Mutexes which is pretty much a requirement to share peripherals and data with interrupt handlers….
125+
2018-10-09 20:51:58 ~japaric min-const-fn is pretty much only constructors and calling other min-const-fns
126+
2018-10-09 20:53:00 @theJPster sounds great. Sorry, I'm slowly catching up here.
127+
2018-10-09 20:53:29 @theJPster So this is this in stable and we're talking about whether to make use of it?
128+
2018-10-09 20:53:39 @therealprof Future stable.
129+
2018-10-09 20:53:55 ~japaric it won't require a feature gate in 1.31{,-beta}
130+
2018-10-09 20:54:12 @theJPster Which crates do we want to make use of this in?
131+
2018-10-09 20:54:19 @therealprof bare-metal
132+
2018-10-09 20:54:44 @therealprof There's a const-fn feature gate which does not work in beta or stable.
133+
2018-10-09 20:54:45 ~japaric heapless too
134+
2018-10-09 20:54:51 @theJPster Oh, I forget that was a crate.
135+
2018-10-09 20:55:33 @therealprof As I said I would keep the feature but turn it into a no-op to make it a non-breaking change.
136+
2018-10-09 20:55:50 @therealprof If people feel strongly we can also completely remove it.
137+
2018-10-09 20:56:49 @theJPster do you mean turn it into a no-op on a rustc where min_const_fn isn't supported?
138+
2018-10-09 20:56:52 jamesmunns Wouldn't that break someone's build, if someone updated their crate but was still using the 1.31-beta, or earlier nightly?
139+
2018-10-09 20:57:31 jamesmunns (oh, japaric already said exactly that, whoops)
140+
2018-10-09 20:57:46 @therealprof Hm, I see what you're getting at.
141+
2018-10-09 20:57:53 ~japaric the only non-breaking change would be to remove #![feature(const_fn)] when the const-fn Cargo feature is enabled
142+
2018-10-09 20:58:08 @therealprof I was kind of assuming ass
143+
2018-10-09 20:58:11 ~japaric users would still need to enable the const-fn Cargo feature to get Mutex::new in const context
144+
2018-10-09 20:58:29 @therealprof …uming that people would use a newer rustc as soon as it is out.
145+
2018-10-09 20:58:42 ~japaric just that they'll be able to use the const-fn Cargo feature in 1.31{,-beta}
146+
2018-10-09 20:58:47 jamesmunns will the crate still build pre-1.31 if a crate has a const fn that they don't use?
147+
2018-10-09 20:59:13 jamesmunns oh
148+
2018-10-09 20:59:14 jamesmunns right
149+
2018-10-09 20:59:16 @therealprof Nope.
150+
2018-10-09 20:59:17 jamesmunns the cfg is still there
151+
2018-10-09 20:59:22 jamesmunns but the feature block isn
152+
2018-10-09 20:59:24 jamesmunns *isnt
153+
2018-10-09 20:59:29 @therealprof Can we version gate that? ;)
154+
2018-10-09 20:59:51 jamesmunns #[cfg(version >= 1.31)]
155+
2018-10-09 21:00:03 ~japaric that doesn't work
156+
2018-10-09 21:00:07 jamesmunns sorry
157+
2018-10-09 21:00:12 jamesmunns tried to write "/s"
158+
2018-10-09 21:00:16 jamesmunns IRC ate it
159+
2018-10-09 21:00:43 ~japaric I mean authors using 1.31 would write crates without cfg-ing on the version and then authors using 1.30 won't be able to compile the crates that depend on cortex-m
160+
2018-10-09 21:01:09 @therealprof Hm?
161+
2018-10-09 21:01:18 jamesmunns I'd say just rev the minor version. I don't think many people are going to care about slightly higher numbers
162+
2018-10-09 21:01:32 @therealprof That's okay for me.
163+
2018-10-09 21:01:44 jamesmunns (not sure if there is any additional overhead when releasing a minor update instead of trivial)
164+
2018-10-09 21:01:55 @therealprof But I think we could make it backwards compatible if we wanted to.
165+
2018-10-09 21:02:31 ~japaric ok, meeting's over. I'll be busy until end of month and super busy until next Wednesday. If there's anything urgent send me an e-mail.

0 commit comments

Comments
 (0)