Re: Memory usage skyrocketed with nim 0.18.0 (in my async tcp service test)
indeed no changes: sudo pmap 28234 28234: ./httpasync0181 561967f7a000124K r-x-- httpasync0181 561968198000 4K r httpasync0181 561968199000 4K rw--- httpasync0181 56196819a000 32K rw--- [ anon ] 561968b41000124K rw--- [ anon ] 7fde9f0e3000 84164K rw--- [ anon ] 7fdea4314000 2172K r-x-- libgc.so.1.0.3 7fdea4533000 4K r libgc.so.1.0.3 7fdea4534000 4K rw--- libgc.so.1.0.3 7fdea4535000264K rw--- [ anon ] 7fdea457c000548K r-x-- ld-musl-x86_64.so.1 7fdea460c000 1472K rw--- [ anon ] 7fdea477c000 4K r--s- localtime 7fdea477f000 8K - [ anon ] 7fdea4781000 84K rw--- [ anon ] 7fdea4796000 8K - [ anon ] 7fdea4798000 84K rw--- [ anon ] 7fdea47ad000 8K - [ anon ] 7fdea47af000340K rw--- [ anon ] 7fdea4804000 4K r ld-musl-x86_64.so.1 7fdea4805000 4K rw--- ld-musl-x86_64.so.1 7fdea4806000 12K rw--- [ anon ] 7fffec4e9000132K rw--- [ pila ] 7fffec5ec000 8K r [ anon ] 7fffec5ee000 8K r-x-- [ anon ] ff60 4K r-x-- [ anon ] total89624K
best practice for writing tests: isMainModule? runnableExamples? tests/ ?
What's the best practice for writing tests? * within module that's being tested (inside isMainModule block)? * within function that's being tested (inside runnableExamples block)? * inside tests/ ? Will alll of these be automatically run by CI and locally with ./koch tests? pros and cons for each? what are the guidelines for when to prefer one or the other? I see all 3 being used in Nim codebase, but consistency is good, this should belong in README.md or in CONTRIBUTING.md ; it could simply be: for tests, do as done proc X in module Y. Even if legacy code doesn't follow these guidelines, new code should follow whatever latest best practices as outlined in some easy to find authoritative source My preference would be to do as D's unittest does, which is to have tests be as local as possible to what's being tested (ie runnableExamples) whenever possible unless there's a good reason not to for a specific thing to test. Does runnableExamples have any drawback? (eg in terms of compile/runtime performance in the case where we are not running tests?)
Re: module name collision across nimble packages
Indeed, you cannot use `std.times.callFun()` as an expression. Nim is not D, I'm sorry.
Re: module name collision across nimble packages
argument for dots: it's nice that declaration matches usage: std.times.callFun() ; whereas we can't use: std/times.CallFun()
Re: Run tests
to compile and run and individual test file in a locally modified nim repo: nim c --lib:lib -r tests/stdlib/talgorithm.nim
Re: NimShooter
> Press A to start > > Actually you have to press Z Ups, sorry, in the retro console mentality the A key is (almost) always the Z key on a pc keyboard... I should fix that > The game looks cool, but is really hardcore. I like that way, for me if a game is not hardcore it has no sense
Re: module name collision across nimble packages
I don't like the dots and think we should deprecate them.
Re: atomic and ref pointer
Well if you `protect` and `dispose` the `ref` properly, you can exchange it between threads. For this to work out it needs to be casted to `pointer` or `ptr`.
Re: NimShooter
> Press A to start > > Actually you have to press Z The game looks cool, but is really hardcore.
Re: how to get unmodified / uninterpreted source code/AST of expression in macro?
That's actually a regular question (pain point?) regarding macros. The current way is to preprocess using an untyped macro, and then dispatch/getAST from a typed macro when you need type resolution. I remember seeing some discussion to ease that but they are pretty buried.
Re: how to skip all the `Hint: foo [Processing]` during compilation?
thanks! embarassed to have asked this... somehow hint[Conf]=off doesn't seem to work (the other ones work); maybe because without that hint it'd be hard to see which configs are being used; in any case would be nice to mention this in nim --advanced entry for hints
Re: how to get unmodified / uninterpreted source code/AST of expression in macro?
> It works when the macro parameter is untyped, thanks @rpowers and @mashingan, using untyped instead of typed works (but I was using typed because of your other answer from [https://forum.nim-lang.org/t/3701/1#23070](https://forum.nim-lang.org/t/3701/1#23070) to allow if x.getTypeInst.typeKind != ntyVoid: to make it work on empty arguments; I guess I need further processing to have both unmodified source AST and allow for void typed arguments)
Re: how to skip all the `Hint: foo [Processing]` during compilation?
You can also put in [any of your nim.cfgs](https://nim-lang.org/docs/nimc.html#compiler-usage-configuration-files) hint[Processing]=off You can also disable other categories of diagnostics besides "Processing" similarly.
Re: how to skip all the `Hint: foo [Processing]` during compilation?
maybe like this? nim c --hints:off themodule
how to skip all the `Hint: foo [Processing]` during compilation?
\--verbosity:0 didn't help also, I don't wanna pipe / filter (not good to rely on shell for that: error prone, can mess up with coloring, not cross platform) ideally there'd be some extensible option, eg: nim c --disp_hints:Processing=false,SuccessX=false,XDeclaredButNotUsed=false,Link=false,Conf=false
Re: Memory usage skyrocketed with nim 0.18.0 (in my async tcp service test)
Actually, the allocator fixes, I think, would only apply to large (>1 GB) memory use, so that may not be relevant here. I'm sorry if I lead you down the wrong path.
Re: How to rewrite this in Nim style?
@Krux02, did you replied the TS code or my tokenize2 ? For removing ref in ref object, I agree. I usually code in value-type instead ref-type, as I only use ref-type when it's usually wrap around costly resource, like socket, heaped memory, file i/o. Agree too with coding by setting pointer for current and latest , that style is discouraged.
Re: how to get unmodified / uninterpreted source code/AST of expression in macro?
This is from tutorial
import macros
macro debug(x: varargs[untyped]): typed =
result = newStmtList()
for s in x:
var slit = s.toStrLit
var stringExpr = slit.repr
echo stringExpr & " okay"
result.add newCall("write", newIdentNode("stdout"), slit)
result.add newCall("write", newIdentNode("stdout"), newStrLitNode(":
"))
result.add newCall("writeLine", newIdentNode("stdout"), s)
var x = 1
proc one(): int = 1
debug(1+1)# prints 1 + 1: 2
debug(x+1)# prints x + 1: 2
debug(one() + 1) # prints one() + 1: 2
What's the problem with toStrLit you have?
Re: how to get unmodified / uninterpreted source code/AST of expression in macro?
It works when the macro parameter is untyped, I guess constant folding happens
during the typing process:
import macros
macro mylog(x:untyped): untyped =
result = newNimNode(nnkStmtList, x)
result.add(newCall("write", newIdentNode("stdout"), toStrLit(x)))
result.add(newCall("write", newIdentNode("stdout"), newStrLitNode(":")))
result.add(newCall("writeLine", newIdentNode("stdout"), x))
var x=1
mylog(x+1) # x + 1:2 (OK but I'd rather want exact source code with
spacing eg: `x+1:2`)
mylog(1+1) # 2:2 (not OK, I want: 1+1: 2)
Re: Memory usage skyrocketed with nim 0.18.0 (in my async tcp service test)
Thanks Araq, right now I am using "clients" to avoid deallocating memory, so that I can get peak memory use. In the future, it will be used to dispatch messages to specific connected clients. This service should also remove clients from that sequence when they disconnect. Sorry, I will try to clean up the code so that the test makes more "sense", this may help me to find out which part is allocating those extra 70MB (Sorry for the long post): sudo pmap 26153 26153: httpasync/src/httpasync018 556fff75a000124K r-x-- httpasync018 556fff978000 4K r httpasync018 556fff979000 4K rw--- httpasync018 556fff97a000 32K rw--- [ anon ] 557000ed2000124K rw--- [ anon ] 7f2806a93000 84164K rw--- [ anon ] 7f280bcc4000 2172K r-x-- libgc.so.1.0.3 7f280bee3000 4K r libgc.so.1.0.3 7f280bee4000 4K rw--- libgc.so.1.0.3 7f280bee5000264K rw--- [ anon ] 7f280bf2c000548K r-x-- ld-musl-x86_64.so.1 7f280bfbc000 1472K rw--- [ anon ] 7f280c12c000 4K r--s- localtime 7f280c12f000 8K - [ anon ] 7f280c131000 84K rw--- [ anon ] 7f280c146000 8K - [ anon ] 7f280c148000 84K rw--- [ anon ] 7f280c15d000 8K - [ anon ] 7f280c15f000340K rw--- [ anon ] 7f280c1b4000 4K r ld-musl-x86_64.so.1 7f280c1b5000 4K rw--- ld-musl-x86_64.so.1 7f280c1b6000 12K rw--- [ anon ] 7ffdf4ad5000132K rw--- [ pila ] 7ffdf4b9c000 8K r [ anon ] 7ffdf4b9e000 8K r-x-- [ anon ] ff60 4K r-x-- [ anon ] total89624K versus 0.17.2 sudo pmap 26094 26094: httpasync/src/httpasync 55b617f37000104K r-x-- httpasync 55b61815 4K r httpasync 55b618151000 4K rw--- httpasync 55b618152000 28K rw--- [ anon ] 55b6194e120K rw--- [ anon ] 7fe5853e4000 11648K rw--- [ anon ] 7fe585f44000 2172K r-x-- libgc.so.1.0.3 7fe586163000 4K r libgc.so.1.0.3 7fe586164000 4K rw--- libgc.so.1.0.3 7fe586165000264K rw--- [ anon ] 7fe5861ac000548K r-x-- ld-musl-x86_64.so.1 7fe586235000 1512K rw--- [ anon ] 7fe5863af000 8K - [ anon ] 7fe5863b1000 84K rw--- [ anon ] 7fe5863c6000 8K - [ anon ] 7fe5863c8000 84K rw--- [ anon ] 7fe5863dd000 8K - [ anon ] 7fe5863df000340K rw--- [ anon ] 7fe586434000 4K r ld-musl-x86_64.so.1 7fe586435000 4K rw--- ld-musl-x86_64.so.1 7fe586436000 12K rw--- [ anon ] 7ffebaff6000132K rw--- [ pila ] 7ffebb124000 8K r [ anon ] 7ffebb126000 8K r-x-- [ anon ] ff60 4K r-x-- [ anon ] total17116K Anyway, I understand that it should work (leak) in a similar way regardless of the gc or nim version used. There is a fixed number of clients that are connecting, so that sequence will grow to 1000 and stop there. I am measuring memory use after all clients (1000) have sent and received 1000 messages, so it should not depend on the speed they are added. jcosborn, I have been doing all tests with "official" docker images (nimlang/nim:0.xx.y-alpine), and nimlang/nim:devel reports version 0.15.3. I have been looking for the original Dockerfile of those images to build an updated devel branch but it seems it is actually generated with nim code in a very smart way I'll put some time on this and report back.
Re: How to count varargs[untyped] inside of template?
@r3d9u11 I don't think you can do that with a template. More information here [https://nim-lang.org/docs/manual.html#templates-varargs-of-untyped](https://nim-lang.org/docs/manual.html#templates-varargs-of-untyped).
