Unable to target armv6 with arm-unknown-linux-gnueabihf C toolchain

gdb backtrace

Starting program: /tmp/_hw 
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1".

Program received signal SIGSEGV, Segmentation fault.
0x004575a6 in ?? ()
(gdb) bt
#0  0x004575a6 in ?? ()
#1  0x004083f2 in _start ()
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
(gdb) disas 0x004575a6
No function contains specified address.
(gdb) disas 0x004083f2
Dump of assembler code for function _start:
   0x004083e0 <+0>:     mov.w   r11, #0
   0x004083e4 <+4>:     mov.w   lr, #0
   0x004083e8 <+8>:
/build/gdb-rWAb4D/gdb-13.1/gdb/displaced-stepping.c:287: internal-error: copy_insn_closure_by_addr: Assertion `buffer.copy_insn_closure.get () != nullptr' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
----- Backtrace -----
0xdf18b ???
0x469edf ???
0x46a163 ???
0x5cb6df ???
0x1734ef ???
0x93f5f ???
0x940a7 ???
0x16fb0f ???
0x171163 ???
0x171763 ???
0x17213b ???
0x1098cb ???
0x10cf97 ???
0x112767 ???
0x42910b ???
0x1f3ad3 ???
0x1f4d47 ???
0x1f3363 ???
---------------------
/build/gdb-rWAb4D/gdb-13.1/gdb/displaced-stepping.c:287: internal-error: copy_insn_closure_by_addr: Assertion `buffer.copy_insn_closure.get () != nullptr' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n)

Tested cargo command/configs

RUSTFLAGS="-C target-cpu=arm1176jz-s -C linker=arm-linux-gnueabihf-gcc-12" cargo build --target arm-unknown-linux-gnueabihf 

RUSTFLAGS="-C target-cpu=generic -C linker=arm-linux-gnueabihf-gcc-12" cargo build --target arm-unknown-linux-gnueabihf 

Target machine information

As I also tested compiling running on the target machine directly (which worked):

arm-linux-gnueabihf-g++-12 --version:

arm-linux-gnueabihf-g++-12 (Raspbian 12.2.0-14+rpi1) 12.2.0
Copyright (C) 2022 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Cross compiling machine information

arm-linux-gnueabihf-g++-12 --version:

arm-linux-gnueabihf-g++-12 (Ubuntu 12.3.0-17ubuntu1) 12.3.0
Copyright (C) 2022 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Further information

• A rust panic-backtrace is not produced
  • I'd wager that this is because it segfaults in start()
• Compiling and running on target machine works as expected (Also with explicit target/target-cpu/linker as with cross)
• Same behavior with debug/release
• Same behavior with 1.88, 1.89, nightly
• Emitted Rust assembly (as compared with --emit=asm) seems to produce identical assembly instructions (sans addresses, identifiers, etc.)
• Arguments to ld (as compared with strace -s99999 -fv -e trace=clone,execve) seem to be identical
• Disassembly with arm-linux-gnueabihf-objdump -d --demangle=rust --no-show-raw-insn --no-addresses produces different output which I cannot compare correctly as I don't know of a tool that can diff the outputs while taking ordering/naming changes into account
• More complex programs sporadically produce SIGILLs that, when debugged with gdb, seem to happen on a bt which seems to be the same instruction when compiled on the target machine - but I could not yet reproduce the SIGILL with a simple program
• Target machine is Raspberry Pi Model B Rev 2 with architecture armhf