@parzivalwolfram@infosec.exchange

ParzivalWolfram @parzivalwolfram@infosec.exchange · Aug 02, 2026

Replying to @whitequark@social.treehouse.systems

@whitequark@social.treehouse.systems if you really want to flex, i believe you can build 3.10 for WinXP, if your code doesn't need newer Python

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Aug 02, 2026

Replying to @xyla@shitpost.trade

@xyla@shitpost.trade the baby badge-in enforcement system is getting wildly out of hand if it's mandatory to buy a TV

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jul 25, 2026

Replying to @ralen@social.ralen.top

@ralen@social.ralen.top Directly after the post you linked to: https://grapheneos.social/@GrapheneOS/116975921189817720

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jul 25, 2026

Replying to @ralen@social.ralen.top

@ralen@social.ralen.top "The POSIX APIs are insecure and outdated..." [citation needed] "...and we don't even want them anyway..." because you're not allowed at the cool kids table? "...but we can't invent a new OS ourselves, so we'll just invent a new kernel with a POSIX shim." no, you won't, because then you won't benefit from free Android security patches and infinitely less people will want to write drivers for your shitass custom kernel.

0

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jul 21, 2026

Replying to @niconiconi@mk.absturztau.be

@niconiconi@mk.absturztau.be you forgot the better option: "the Linux driver was rewritten from scratch and is now infinitely better but it's not 80 column compliant so it's forever stuck as a random git repo no one will ever find, and reliant on dkms to use it."

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jul 07, 2026

Replying to @parzivalwolfram@infosec.exchange

of course I fix time.h so everything's not locked behind a million ifdefs for K&R support and the like, and gcc-6.5.0 starts segfaulting on libquadmath mid-build (and the --disable-libquadmath flag for configure doesn't work, because why would it?)

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jul 04, 2026

Replying to @parzivalwolfram@infosec.exchange

this is the sixth GCC version i've seen that requires you to rebuild libstdc++-v3 after building it proper, which means you can't do a normal bootstrap and call it good. This has applied even on Linux. What the fuck is wrong with GCC most of the time that makes it this hard to build?

0

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jul 04, 2026

Replying to @parzivalwolfram@infosec.exchange

i have skipped over half the compiler BUT IT FUCKING WORKS now i just have to get libstdc++ and libatomic to build lol

0

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 09, 2026

Replying to @parzivalwolfram@infosec.exchange

"how is this un-debuggable" the default debugger, adb, doesn't catch subprocesses' segfaults, can't pass argv to things, and generally sucks dick. I have yet to find a version of gdb that doesn't build fine but immediately segfault on running a process.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 09, 2026

Replying to @parzivalwolfram@infosec.exchange

yeah, there's the undebuggable segfault. it took a while, it got most of the way through libstdc++-v3

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 09, 2026

Replying to @parzivalwolfram@infosec.exchange

It won't build GCC 4.9... but it WILL build stage1 of GCC 6.5.0, which was a complete "this will be funny for the thread lmao" move that seems to have paid off????? this is gonna be crushing later when it doesn't work in like stage3 and can't easily be fixed

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 07, 2026

Replying to @parzivalwolfram@infosec.exchange

well that gives me binutils-2.25.1 before stdint.h is a hard requirement so there's that, i guess. Let's see if that'll build gcc 4.9 with its weird-ass C++ linking behavior.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 07, 2026

Replying to @parzivalwolfram@infosec.exchange

ah yes, i'm chasing doen a bug where all versions of binutils after 2.17 claim all files are truncated, i accidentally turn off my custom headers for one pass of binutils-2.20.1 and the bug disappears. why is ONLY binutils affected by this and something like gcc 4.5.4 or python 2.7.16 or SDL2 2.30.0 isn't?

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 05, 2026

Replying to @parzivalwolfram@infosec.exchange

oh wow, having a working linker... makes things actually work! who would've thought?

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 05, 2026

Replying to @parzivalwolfram@infosec.exchange

lol. lmao, even

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jun 05, 2026

Replying to @parzivalwolfram@infosec.exchange

so it turns out at some point in 2003 they fixed the bug and put out a patch for this. for those googling later, it's in the Recommended patch clusters from 2006-ish. (patch ID is 106951-25 for x86, 106950-24 for SPARC.)

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · May 07, 2026

Replying to @xyla@shitpost.trade

@xyla@shitpost.trade nice try, but you'll NEVER know how old my kernel is!

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · May 02, 2026

Replying to @parzivalwolfram@infosec.exchange

Yup, looks like an ld.so.1 bug in Solaris 7. https://gcc.gnu.org/pipermail/gcc-bugs/2002-October/064911.html It is interesting that they have working gdb, though, i'll have to see if 5.0 builds, as it'd be infinitely easier to use than fucking adb.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · May 02, 2026

Replying to @parzivalwolfram@infosec.exchange

Ah yes, we're now segfaulting in, uh... ...ah, fuck, that's init code isn't it? we're not even *in* the ELF yet, are we?

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Apr 24, 2026

Replying to @parzivalwolfram@infosec.exchange

see, this is an issue because sol7 defines both long long and int64/uint64 as union(int32*2,double) "for alignment purposes", as you're not meant to actually try to USE them, and fseeko64 is long long, so it's completely valid. After staring into adb (no, not the Android kind) for way too long, it appears what the SIGILLs and SIGSEGVs are from is it's encoding constants into ELF files as 64 bits, inline, as i've noticed almost all of the SIGILLs happen directly after a register store, so something like "mov r32,imm32" and then an illegal 4-byte "instruction." Sounds a lot like it's emitting a 64-bit imm32, which would explain why it's so flaky: the first 32 bits might make it into the register, and then the second 32 fucking don't and might or might not be a valid opcode (the second 32 bits are usually not \x00, so i'm assuming something's also going wrong with the offsets, as no normal ELF is going to be a few terabytes into the address space on a fucking 486.)

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Apr 24, 2026

Replying to @parzivalwolfram@infosec.exchange

No, no it's so much worse than that: libbfd only cares if you have fseeko64 for 64-bit host support. which we do, but it's a lie, and it doesn't just use FILE_OFFSET_BITS in that case, it uses a whole different half of the fucking makefile and undefines FILE_OFFSET_BITS entirely. Oh boy.

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Apr 18, 2026

Replying to @parzivalwolfram@infosec.exchange

Ah! Once again, vbox bites me in the ass: most things trust CPUID, *except* libbfd, which seems to build assembly files to test your processor's real capabilities. This might be where all these SIGILLs are from: vbox is masking CPUID, but libbfd is still picking up HOST_BFD_64_BIT on a 32-bit OS.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Apr 07, 2026

Replying to @parzivalwolfram@infosec.exchange

tragic: GCC fails to link to any library containing symbol "operator new[](void *)", is asked to leave build folder. Onlookers added context at the scene: "What the fuck is GCC doing, that's not how you link to a function. I don't think symbols can even have spaces in their names?"

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Feb 25, 2026

Replying to @parzivalwolfram@infosec.exchange

ah, yes, now CMake forgets that "file.c" is a C file, and "file2.cxx" is a CXX file. How did I fuck this up this badly, exactly?

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Feb 15, 2026

Replying to @parzivalwolfram@infosec.exchange

Again? Again. Earlier, this time. This worked an hour ago, and now by the magic of "here's off_t" we've toasted wchar.h, which has never once needed a patch before. what

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Feb 11, 2026

Replying to @parzivalwolfram@infosec.exchange

yes thank you gcc, "a != a"

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Feb 10, 2026

Replying to @parzivalwolfram@infosec.exchange

we have stdatomic. thank you to Jens Gustedt, who put together a libatomic I could understand enough to get working. https://gitlab.inria.fr/gustedt/stdatomic

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 29, 2026

Replying to @parzivalwolfram@infosec.exchange

are build success reports for this platform even helpful if I've had to patch both the program AND the tests? something like patching around MAP_ANON being missing and not being confident at all that it's correct, but it is working? I probably shouldn't submit those, right?

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 27, 2026

Replying to @parzivalwolfram@infosec.exchange

uh oh perl, did you ignore the -ggdb flag?

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 25, 2026

Replying to @parzivalwolfram@infosec.exchange

The answer is no. perl-5.38.0 is occasionally installing itself to $PREFIX/$PREFIX/ instead of $PREFIX/ and this is only happening after moving it up to before gcc 4.2.4 instead of after (the thing responsible for setting this data is a fucking *shell script* so why would that matter?)

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 24, 2026

Replying to @parzivalwolfram@infosec.exchange

Perl, can you stop being a dick for five seconds? I have the build system set up specifically for a reason, obey my settings!

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 24, 2026

Replying to @parzivalwolfram@infosec.exchange

Solaris 7 Pro Tip: after installing XFree86 and getting it set up, use the following one-liner to get it to take over the display server: echo ":0 Local local_uid@console root /usr/X11R6/bin/X" > /usr/dt/config/Xservers && reboot

(nowhere in the docs is this clear at all, and the docs lie in general.)

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 23, 2026

Replying to @parzivalwolfram@infosec.exchange

it appears it's time once again.

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 23, 2026

Replying to @parzivalwolfram@infosec.exchange

i can't even build binutils-2.13.2.1 anymore, which came before GCC in general earlier. i hate this shit dude it's rotting BEFORE MY EYES

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 19, 2026

Replying to @parzivalwolfram@infosec.exchange

I've had friends suggest things during the duration of this project, some suggestions are helpful and some are very much not: "Rebuild your scripts, they're ass and hard to maintain" - helpful, broke shit but needed done. "You saw that one Unix book in that store, right? Buy it if it has the CD, i'll just give you the money." - Suspicious, but helpful, contained a lot of good hints and also seemingly lost versions of a lot of GNU software that I can bootstrap with suncc. archive.org listing soon. "Just port glibc!" - trust me buddy, i've tried, it's not as easy as "add sysdefs lmao" like they make it out to be and they removed the ability to port it in like 2014 "Just port GCC !" - god no could you imagine how fucking hard it would be to deal with the macro hell in there? "Just use modern Linux" - not helpful, that's no fun. "Use rust lmao" - not helpful, even taking it seriously I'd have to add full C98/C11/C++11 at minimum to build the vendored LLVM and if I could do that the issue would be over anyways. "Build your own compiler equivalent to GCC !" - i'm not smart enough to not fuck up the system headers, and you want me to do WHAT

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 17, 2026

Replying to @parzivalwolfram@infosec.exchange

gdb be like "yeah we need to make this library. oh but we can't actually because the un-built library is missing sorry"

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 14, 2026

Replying to @parzivalwolfram@infosec.exchange

ah yes, we can BUILD binutils-2.20.1, but if you EVER try to install it it eats shit on INSTALLING a static library. literally how

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 10, 2026

Replying to @parzivalwolfram@infosec.exchange

this legitimately instantly fixed a hundred things because several GNU makefiles reset your PATH at some point and if you're trying to avoid default linker/etc, suddenly resetting the path tends to revert back to them, even with "LD=/very/specific/full/path/to/ld" and such.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 10, 2026

Replying to @parzivalwolfram@infosec.exchange

$ cat sys-patch.sh #Go To Jail For Your Linker Crimes mkdir -p /usr/jail mv /usr/ccs/bin/as /usr/jail/as.orig mv /usr/ccs/bin/ld /usr/jail/ld.orig

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 10, 2026

Replying to @parzivalwolfram@infosec.exchange

challenge to the gnu foundation: do literally anything other than override explicit instructions from the user in thirty separate places

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 09, 2026

Replying to @parzivalwolfram@infosec.exchange

i've gone ahead and rewritten 90% of my build system, and now I can build binutils-2.15 (kind of) but now gcc-3.2.3 is inconsistent on whether ld will SIGSEGV or not, and if it doesn't, tar-1.25 can't be built because it loses track of its own internal includes (alloca.h, part of libgnu), which it only tells me about at link time.

Nothing has changed in the env itself, I've got all kinds of debug output to make sure of that.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Jan 09, 2026

Replying to @parzivalwolfram@infosec.exchange

man /opt/SUNWspro/bin/cc "yes we have similar -O flags to GCC but no -OS or -Os" ok, that's reasonable...

-O1 "WARNING: Passing '-1' to linker!"

so that was a fucking lie

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 29, 2025

Replying to @parzivalwolfram@infosec.exchange

"illegal instruction" oh boy, gcc 4.5.4 bootstrap isn't working again! how did i fix this last time?

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 25, 2025

Replying to @parzivalwolfram@infosec.exchange

The issue is more than this one bug, if you're dealing with this bug, this is how you fix it, as it starts at the very beginning of the non-Sun compiler stack:

Don't use tgcware or sunfreeware packages. They're all poisoned in this way, and it causes issues compiling just about everything later down the line. Start with SUNWspro.
Build binutils 2.13 before gcc 2.x, which is required for 3.x or higher anyway. SUNWspro will just barely suffice for this on Solaris 7.
Build gcc 2.95.3, ENFORCE THE GNU BINUTILS. The docs say to use Solaris as+ld, this is a trap. Don't fall for it! Never use them unless there's strictly no other option.
Build binutils 2.15, then gcc 3.2.3 (do not bootstrap.) Because gcc 2.95.3 isn't poisoned by the system linker, you'll be OK.
As gcc 3.2.3 cannot be bootstrapped with 2.95.3 (despite that being the opposite of every doc in the universe about gcc being forced to bootstrap, gcc 2.95.3 is not capable of doing this) remove the GCC 2.x specific include dir in your prefix (3.2.3 won't do this itself) and immediately rebuild gcc 3.2.3 (with bootstrap-lean) to remove errant C++ behavior for later.
Build binutils 2.17. 2.15 will build gcc 4.x and beyond, but this breaks libstdc++ permanently as it'll version symbols that do not exist. If "floor@@SUNW_1.1" symbol refs exist in a binary, you've hit this bug.
Build gcc 4.2.4, disable bootstrap. (It will bootstrap anyway, but if you try to enable it, it'll break. This occurs with make 3.7 through 4.4, so it's a logic bug in gcc's autotools set.) As before, remove all 3.2.3-specific include directories (and $PREFIX/i386-pc-solaris2.7/assert.h, due to a CFLAGS bug in 4.2.4 through 4.6, and probably much further) and then immediately rebuild it with itself. (Again, disable bootstrap, it will anyway.)

At this point, if libstdc++.so has no references to __eprintf or floor@@SUNW_1.1, you're in the clear. More testing is needed to see if this persists through 4.6.x, the last versions with any measure of sol7 support, but i'll keep people posted.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 15, 2025

Replying to @parzivalwolfram@infosec.exchange

figured out why libgnu.a becomes trashed due to __eprintf being referenced but hidden: binutils 2.15 poisons gcc 3.2, which then cascades ALL the way up to gcc-4.3.6 before problems actually show up, when symbols support being hidden suddenly. There's a bug and everything, which was wontfixed and forgotten about 20 FUCKING YEARS AGO, and 5 years before the bug actually trips in gcc on sol7! https://gcc.gnu.org/bugzilla/show_bug.cgi?id=19461 motherfuckers, all of them

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 15, 2025

Replying to @parzivalwolfram@infosec.exchange

with some help from a friend libffi-3.5.2 is BUILT but not TESTED because i'm not doing fucking dejagnu, it's so much extra shit to build to test a couple packages... let's see if python thinks it works.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 14, 2025

Replying to @parzivalwolfram@infosec.exchange

Ah, yes: dd from coreutils-9.7 DOES WORK... unless status is enabled, at which point it will segfault as it's taking "jd" from... the formatted strings, I think? and just taking that literally, and then segfaulting shortly after when trying to compute numbers against string "jd", which I'm surprised doesn't work, as I figured you COULD, in fact, just do calculations against a string... The fun part is that it works fine otherwise, so I've been compiling things normally and correctly with a busted dd the whole time.

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 14, 2025

Replying to @parzivalwolfram@infosec.exchange

you know a package is completely hosed when you have to modify like twelve files to get CFLAGS to be respected, and it's especially bad when the CFLAGS need to be respected because they've somehow fucked up includes enough to require explicitly passing "-I/usr/include" wouldn't it be fucking hilarious if that package was gcc-4.3?

1

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 07, 2025

Replying to @parzivalwolfram@infosec.exchange

how the fuck do you debug assembler errors from GCC, assuming a .S input? are the aline listings before or after preprocessing? the gcc and as docs don't specify, despite furious grepping

0

ParzivalWolfram @parzivalwolfram@infosec.exchange · Dec 06, 2025

Replying to @parzivalwolfram@infosec.exchange

what the fuck, gcc?

0

ParzivalWolfram

Posts

Remote instance

Media