Tail calls passing structs had been broken since forever, and it made me mad enough to write C++.
It's a little tricky because sometimes it requires shuffling values around on the stack and you have to make sure to not overwrite values you'll need later. This was fixed in the arm backend a while ago, and I was able to port that approach to the x86 backend. The plan is to include this in LLVM 22.
Together with some other changes it seems like guaranteed tail calls in rust might actually, finally happen.
Po emigracji z pol.social wypada się przedstawić, toteż czynię tu #introduction.
Jam jest Jezus Michał (czyt. "Jezus, Michał!"), pan #kot twój po trzykroć, i #mikol. Mieszkam w okolicy miasta #Poznań, gdzie #kolej z reguły funkcjonuje, więc wyznaję zasadę "każdy powinien mieć sieciówkę" i filozofię #BezSamochodu. Na koncie mam 38% polskiej sieci normalnotorowej (wg. viaduct.world), ale pracuję nad tym. Przy okazji włóczę się po lasach, i uzupełniam #OpenStreetMap.
W "wolnym" czasie męczę #WolneOprogramowanie. W roku 2010 wpadłem w pułapkę i dołączyłem do współtwórców dystrybucji #Gentoo. Tam zajmuję się wszystkim po trochu, aczkolwiek główne tematy to #Python, #LLVM (za jakie grzechy…), #Xfce. Czekam na #JihadButleriański.
Obok tego uprawiam retorykę na #AntyKapitalizm. Cenię #SłowaNaOpak i konkretny #suchar. Od kilku lat dręczy mnie #cukrzyca, mam też stwierdzone spektrum autyzmu [#autyzm]. A, zapomniałbym — "Nie potrafię ciekawie spędzać czasu" (nie, ja nigdy nie zapominam).
If an individual work is unprotected by copyright law in the
United States and you are located in the United States, we do not
claim a right to prevent you from copying, distributing, performing,
displaying or creating derivative works based on the work as long as
all references to Project Gutenberg are removed.
using the license agreement (which only has weight in copyright regimes like the us) specifically to require distributors to remove references to project gutenberg, in order to avoid a paper trail identifying project gutenberg
# Indicates whether the LLVM testsuite is enabled in the build or not. Does<br></br># not execute the tests as part of the build as part of x.py build et al,<br></br># just makes it possible to do `ninja check-llvm` in the staged LLVM build<br></br># directory when doing LLVM development as part of Rust development.<br></br>#llvm.tests = false<br></br>
did no one question "hey why do we need to set these things here in the first place if it doesn't matter and we can just delegate to llvm's existing build system?"? idk
Also I am curious about #FreeBSD with #ports and #poudriere, how do you guys manage #Firefox and possibly #LibreOffiice? It took me~5h to compile #LLVM default flavor on my laptop, I would just assume giants I listed above will take more than 10 hours? I still remember old days when I was using #Gentoo#Linux and whenever there was updates for them I had to keep my PC on overnight...But nowadays #Firefox seems to update more frequently, I dare not to compile few times a month.
the gist is that I'd like to do substantial upgrades to both the hardware and software side of the online Alive2 web site that the #llvm community uses:
putting this online in case it catches the interest of anyone working for a company that might have some $$ for this sort of thing, since the situation at NSF is grim
Linking scopesrt
/usr/bin/ld: /home/lritter/devel/duangle/scopes/clang/lib/libLLVMX86Disassembler.a: error adding symbols: file format not recognized
clang++: error: linker command failed with exit code 1 (use -v to see invocation)
ok. nobody on the #llvm channel replies. i can find no relevant results on google. this is a complete mystery. either they did something wrong or i did. only thing i can do is try more releases and find the exact point where the .a files got messed up.
Thanks to Florian's endless follow-up work, the TypeSanitizer finally landed, and you should be able to do clang -fsanitize=type (should be live on Godbolt tomorrow)! 🎉 #LLVM
Had a bunch of thoughts about the recent safety stuff, way more than fit in social media post... Blog post story time! (It's a bit of a ramble, sorry about that...)
This week is chock full of conference activity for Igalians! Several of us will be attending the WebKit Contributors Meeting in Cupertino, California, which starts today.
Today is also the start of the #RISCV Summit US in Santa Clara, California, where tomorrow,
@asb will give a talk about the current state of #LLVM development.
trying to insert a layer above s-exprs where we just expose scopes' token stream (similar to how rust does it for syntax macros), which is slightly more complex since we need to generate tokens for indentation changes; and we can't really expose auto-wrapping of single arguments and left-wrapping ; operators since these require arbitrary read-ahead to produce a token in advance, which would defeat the purpose.
but i think this is a good way forward since it's representation independent.
@breakin i want it to sound like a standard as much as possible
(fun fact: i used to be the only creative guy in the #llvm IRC channel; so much so that the project lead approached me to choose a new name for them. after a week i told him i thought that LLVM was already perfect, and artsy names would make it sound like it wasn't a serious project, and that was it ;-)