Executing binary with link loader SIGSEVs about nix-ld

[Mic92]

Fixed in c7149c0

[magneticflux-]

I'm running into the same issue with the latest nix-ld:

❯ strace /nix/store/386r5hy238s7djbpdpijxlj2f8kfz7cw-nix-ld-1.2.2/libexec/nix-ld
execve("/nix/store/386r5hy238s7djbpdpijxlj2f8kfz7cw-nix-ld-1.2.2/libexec/nix-ld", ["/nix/store/386r5hy238s7djbpdpijx"...], 0x7ffed4cf6e20 /* 59 vars */) = 0
openat(AT_FDCWD, "/run/current-system/sw/share/nix-ld/lib/ld.so", O_RDONLY) = 3
read(3, "\177ELF\2\1\1\3\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0@\252\1\0\0\0\0\0"..., 64) = 64
mmap(NULL, 680, PROT_READ, MAP_PRIVATE, 3, 0) = 0x7f3f416c3000
mmap(NULL, 217088, PROT_READ, MAP_PRIVATE, 3, 0) = 0x7f3f4168e000
mmap(0x7f3f4168f000, 155648, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_FIXED, 3, 0x1000) = 0x7f3f4168f000
mmap(0x7f3f416b5000, 40960, PROT_READ, MAP_PRIVATE|MAP_FIXED, 3, 0x27000) = 0x7f3f416b5000
mmap(0x7f3f416bf000, 16384, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x31000) = 0x7f3f416bf000
munmap(0x7f3f416c3000, 680)             = 0
close(3)                                = 0
execve("/cur/current-system/sw/share/nix-ld/lib/ld.so", ["/nix/store/386r5hy238s7djbpdpijx"...], 0x7ffd5f080458 /* 59 vars */) = -1 ENOENT (No such file or directory)
brk(NULL)                               = 0x555556055000
arch_prctl(0x3001 /* ARCH_??? */, 0x7ffd5f0802f0) = -1 EINVAL (Invalid argument)
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x238} ---
+++ killed by SIGSEGV (core dumped) +++
fish: Job 1, 'strace /nix/store/386r5hy238s7d…' terminated by signal SIGSEGV (Address boundary error)

Was the bug reintroduced, or am I doing something wrong? It works if I use bash, and then even if I open fish again inside that, but the original fish shell causes a SIGSEGV.

Valgrind:

valgrind /lib64/ld-linux-x86-64.so.2
==113484== Memcheck, a memory error detector
==113484== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.
==113484== Using Valgrind-3.21.0 and LibVEX; rerun with -h for copyright info
==113484== Command: /lib64/ld-linux-x86-64.so.2
==113484==
==113484== Invalid read of size 4
==113484==    at 0x4807B07: _dl_process_pt_gnu_property (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481D60E: dl_main (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481B472: _dl_sysdep_start (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481CB9B: _dl_start (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481BA47: ??? (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x3D485441505F5951: ???
==113484==    by 0x7275632F6E75722E: ???
==113484==    by 0x7379732D746E6571: ???
==113484==    by 0x732F77732F6D6573: ???
==113484==    by 0x78696E2F65726167: ???
==113484==    by 0x62696C2F646C2C: ???
==113484==    by 0x2F3D444C5F58494D: ???
==113484==  Address 0x238 is not stack'd, malloc'd or (recently) free'd
==113484==
==113484==
==113484== Process terminating with default action of signal 11 (SIGSEGV): dumping core
==113484==  Access not within mapped region at address 0x238
==113484==    at 0x4807B07: _dl_process_pt_gnu_property (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481D60E: dl_main (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481B472: _dl_sysdep_start (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481CB9B: _dl_start (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x481BA47: ??? (in /nix/store/905gkx2q1pswixwmi1qfhfl6mik3f22l-glibc-2.37-8/lib/ld-linux-x86-64.so.2)
==113484==    by 0x3D485441505F5951: ???
==113484==    by 0x7275632F6E75722E: ???
==113484==    by 0x7379732D746E6571: ???
==113484==    by 0x732F77732F6D6573: ???
==113484==    by 0x78696E2F65726167: ???
==113484==    by 0x62696C2F646C2C: ???
==113484==    by 0x2F3D444C5F58494D: ???
==113484==  If you believe this happened as a result of a stack
==113484==  overflow in your program's main thread (unlikely but
==113484==  possible), you can try to increase the size of the
==113484==  main thread stack using the --main-stacksize= flag.
==113484==  The main thread stack size used in this run was 8388608.
==113484==
==113484== HEAP SUMMARY:
==113484==     in use at exit: 0 bytes in 0 blocks
==113484==   total heap usage: 0 allocs, 0 frees, 0 bytes allocated
==113484==
==113484== All heap blocks were freed -- no leaks are possible
==113484==
==113484== For lists of detected and suppressed errors, rerun with: -s
==113484== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 0 from 0)
fish: Job 1, 'valgrind /lib64/ld-linux-x86-64…' terminated by signal SIGSEGV (Address boundary error)