logs archiveBotHelp.net / Freenode / #2f30 / 2015 / August / 14 / 4
Evil_Bob
heh
stateless
haha
didn't know that
Evil_Bob
(linux mmap)
linux seems to implement some extra hints and flags to mmap, while OpenBSD doesnt
(tested on openbsd though)
stateless
yes
Evil_Bob, another possibility is to unlink the file, if processes are serialized
process A) mmap private, process B) unlink, truncate, process C) unlink, fill
assuming B and C cannot run at the same time
this might work
Evil_Bob
ah
stateless
the file before B) or C) will be retained until you call munmap() in process A)
so the truncate and fill operations later on are modifying a new file
unrelated to the mmap
atmc
hm
what project need work atm?
will code c/go for jiggles
or catpicz
stateless
Evil_Bob, http://www.mail-archive.com/linux-btrfs@vger.kernel.org/msg01647.html
had to send this to some people at work
:P
atmc, we need a correct and efficient code generator for x86-64 in scc
you have until Sunday eve
atmc
code generator? :D
what you think about void linux btw?
and void vs alpine
Evil_Bob
stateless: "But real gcc
developers really thought that it makes sense, because the standard allows
it, and it gives the compiler the maximal freedom - because it can now do
things that are CLEARLY NONSENSICAL."
GNU in a nutshell
k0ga
the code generator is too much, but if you can write tests,, this can be very helpful now
stateless
Evil_Bob, hahaha
except that GNU does nonsensical things not necessary in the name of the standard
Evil_Bob
lol true
k0ga
stateless: very good the link
Evil_Bob
stateless: wtf, live kernel debugging http://www.flabber.nl/artikel/patient-in-utrecht-medisch-centrum-zingt-opera-tijdens-hersenoperatie-45425
at 2:40 he almost kernel panics
(from a dutch hospital in Utrecht)
heh this is awesome
atmc
k0ga: eh tests, like tdd tests? :D
k0ga
I developed the front end fast, to have something that can be tested
atmc
that
seems
tedious
hell
k0ga
it is mature enough to begin making tests
atmc: no, it is only thinking things that the compiler should detect
fix the bugs is my bussines ;)
atmc
i mean even writing the tests
but the front end for testing is cool tbh
its just a few line bash script
and its in fact a full testing framework
k0ga
atmc: check it again. I have pushed the last version
stateless
k0ga, http://sprunge.us/igOM
what's the deal with this code? :
test.c:9: error: 'fn' initialized and declared extern
k0ga
stateless: initializers are not done
stateless
ok
it works if i move the assignment down
dami0
that's because then it doesn't run, no?
stateless
dami0, no it is because the compiler doesn't support initialization at that point
yet
dami0
oh, i thought that the C spec didn't really specify that code had to be run after a return
stateless
k0ga, nice that int ar[5]; 1[ar] = bla works :P
dami0, what return?
that code just initializes the function pointer, it doesn't call it
k0ga
stateless: I have pushed a small change that makes cc1 accepts initializers, but it doesn't verify the correctnes and doesn't emit anything
dami0
there's a return in there, i thought if you put code after the function's return value, it's not guaranteed to execute
k0ga
you can test you first version now ;)
dami0
are you guys making your own compiler?
k0ga
dami0: yes
dami0
cool
stateless
dami0, there's no code after return
dami0
>stateless@ : it works if i move the assignment down
stateless
obviously before the return
void (*fn)(void);
fn = foo;
return 0;
dami0
ah
k0ga
stateless: int ar[5]; 1[ar] = - works because it is a consequence of how the parser works
stateless
yeah nice
k0ga, ready? :P
k0ga
for what?
stateless
http://sprunge.us/ULIP
try this :P
k0ga
you discovered an error :P
stateless
thank u
where is my money.
k0ga
can you push a test for it?
stateless
sure
k0ga, I won't include sample output
k0ga
no
stateless
cuz I don't know what it looks like
k0ga
because the output is complex
stateless
ok
so just the .c renamed to test003.c?
k0ga
I have to verify the output and see what is the correct value
it is not easy even for me
« prev 1 2 3 4 5 next »