RustiCL seems to be crashing on a std::string constructor. Detected it when running a custom Darktable build, but reproduceable by simply using the `clinfo` tool. In clinfo, the crash backtrace looks like this: ``` Thread 1 "clinfo" received signal SIGSEGV, Segmentation fault. std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_construct<true> (this=0x7fffffffaf90, __str=0x555555 <error: Cannot access memory at address 0x555555>, __n=0) at /usr/include/c++/15/bits/basic_string.tcc:298 warning: Source file is more recent than executable. 298 this->_S_copy(_M_data(), __str, __n + _Terminated); (gdb) bt #0 std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_construct<true> (this=0x7fffffffaf90, __str=0x555555 <error: Cannot access memory at address 0x555555>, __n=0) at /usr/include/c++/15/bits/basic_string.tcc:298 #1 0x00007ffff76b8abf in std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string (this=0x7fffffffaf90, __str="") at /usr/include/c++/15/bits/basic_string.h:238 #2 spvtools::LinkerOptions::GetFnVarArchitecturesCsv[abi:cxx11]() const (this=<optimized out>, this=<optimized out>) at /usr/src/debug/spirv-tools-2025.4-1.fc43.x86_64/include/spirv-tools/linker.hpp:76 #3 spvtools::Link (context=..., binaries=binaries@entry=0x555555579b60, binary_sizes=binary_sizes@entry=0x555555579720, num_binaries=<optimized out>, linked_binary=linked_binary@entry=0x7fffffffb730, options=...) at /usr/src/debug/spirv-tools-2025.4-1.fc43.x86_64/source/link/linker.cpp:858 #4 0x00007ffff76bc671 in spvtools::Link (context=..., binaries=..., linked_binary=0x7fffffffb730, options=...) at /usr/src/debug/spirv-tools-2025.4-1.fc43.x86_64/source/link/linker.cpp:820 #5 0x00007ffee638df48 in clc_link_spirv_binaries (args=0x7fffffffb870, logger=0x7fffffffb858, out_spirv=out_spirv@entry=0x7fffffffb880) at ../src/compiler/clc/clc_helpers.cpp:1337 #6 0x00007ffee638aec5 in clc_link_spirv (args=<optimized out>, logger=<optimized out>, out_spirv=0x7fffffffb880) at ../src/compiler/clc/clc.c:146 #7 0x00007ffee5a9ed2d in mesa_rust::compiler::clc::spirv::SPIRVBin::link (spirvs=..., library=false) at ../src/gallium/frontends/rusticl/mesa/compiler/clc/spirv.rs:198 #8 0x00007ffee5a6a9ac in rusticl::core::program::Program::build (self=0x555555572020, devs=..., options=...) at ../src/gallium/frontends/rusticl/core/program.rs:580 #9 rusticl::api::program::build_program (program=0x555555572070, num_devices=<optimized out>, device_list=<optimized out>, options=<optimized out>, pfn_notify=<error reading variable: Cannot access memory at address 0x0>, user_data=<optimized out>) at ../src/gallium/frontends/rusticl/api/program.rs:340 #10 rusticl::api::program::clBuildProgram (program=0x555555572070, num_devices=<optimized out>, device_list=<optimized out>, options=<optimized out>, pfn_notify=..., user_data=0x0) at ../src/gallium/frontends/rusticl/api/program.rs:316 #11 0x0000555555560434 in getWGsizes (wgm_sz=1, output=<optimized out>, ret=<optimized out>, loc=0x7fffffffbba0, wgm=0x7fffffffbac8) at src/clinfo.c:2019 #12 device_info_wg (ret=0x7fffffffbbe0, loc=0x7fffffffbba0, chk=<optimized out>, output=<optimized out>) at src/clinfo.c:2076 #13 0x0000555555560c58 in printDeviceInfo (dev=dev@entry=0x555558e754e0, plist=plist@entry=0x7fffffffc260, p=p@entry=1, param_whitelist=<optimized out>, param_whitelist@entry=0x0, output=output@entry=0x7fffffffc800) at src/clinfo.c:3709 #14 0x000055555556233c in printPlatformDevices (plist=0x7fffffffc260, p=1, device=0x55555576f560, ndevs=<optimized out>, str=0x7fffffffc2b0, output=0x7fffffffc800, these_are_offline=0) at src/clinfo.c:3993 #15 0x0000555555555ecd in showDevices (plist=0x7fffffffc260, output=0x7fffffffc800) at src/clinfo.c:4051 #16 main (argc=<optimized out>, argv=<optimized out>) at src/clinfo.c:4849 `` Reproducible: Always Steps to Reproduce: clinfo -d 1:0 Actual Results: ➜ ~ clinfo -d 1:0 Platform Name rusticl Device Name AMD Radeon RX 7900 XT (radeonsi, navi31, LLVM 21.1.5, DRM 3.64, 6.17.12-300.fc43.x86_64) Device Vendor AMD Device Vendor ID 0x1002 Device Version OpenCL 3.0 Device UUID 00000000-0b00-0000-0000-000000000000 Driver UUID 414d442d-4d45-5341-2d44-525600000000 Valid Device LUID No Device LUID 0000-000000000000 Device Node Mask 0 Device Numeric Version 0xc00000 (3.0.0) Driver Version 25.2.7 Device OpenCL C Version OpenCL C 1.2 Device OpenCL C Numeric Version 0x402000 (1.2.0) Device OpenCL C all versions OpenCL C 0xc00000 (3.0.0) OpenCL C 0x402000 (1.2.0) OpenCL C 0x401000 (1.1.0) OpenCL C 0x400000 (1.0.0) Device OpenCL C features __opencl_c_integer_dot_product_input_4x8bit 0x800000 (2.0.0) __opencl_c_integer_dot_product_input_4x8bit_packed 0x800000 (2.0.0) __opencl_c_int64 0x400000 (1.0.0) __opencl_c_kernel_clock_scope_device 0x400000 (1.0.0) __opencl_c_kernel_clock_scope_sub_group 0x400000 (1.0.0) __opencl_c_images 0x400000 (1.0.0) __opencl_c_read_write_images 0x400000 (1.0.0) __opencl_c_3d_image_writes 0x400000 (1.0.0) __opencl_c_ext_image_unorm_int_2_101010 0x400000 (1.0.0) __opencl_c_subgroups 0x400000 (1.0.0) Latest conformance test passed v0000-01-01-00 Device Type GPU Device PCI bus info (KHR) PCI-E, 0000:0b:00.0 Device Profile EMBEDDED_PROFILE Device Available Yes Compiler Available Yes Linker Available Yes Max compute units 84 Max clock frequency 2175MHz Device Partition (core) Max number of sub-devices 0 Supported partition types None Supported affinity domains (n/a) Max work item dimensions 3 Max work item sizes 1024x1024x1024 Max work group size 1024 Preferred work group size multiple (device) 64 [1] 239982 segmentation fault (core dumped) clinfo -d 1:0 Expected Results: clinfo runs without crashing Additional Information: I have an AMD Radeon 79000 XT: ``` ➜ ~ lspci | grep -i "vga.*navi" 0b:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Navi 31 [Radeon RX 7900 XT/7900 XTX/7900 GRE/7900M] (rev cc) ```
`clinfo` does not fail when using the ROCM platform on this same machine: ``` ➜ ~ clinfo -l Platform #0: AMD Accelerated Parallel Processing `-- Device #0: gfx1100 Platform #1: rusticl `-- Device #0: AMD Radeon RX 7900 XT (radeonsi, navi31, LLVM 21.1.5, DRM 3.64, 6.17.12-300.fc43.x86_64) ``` ``` ➜ ~ rpm -qf $(which clinfo) clinfo-3.0.25.02.14-1.fc43.x86_64 `` ```➜ ~ clinfo -d 0:0 Platform Name AMD Accelerated Parallel Processing Device Name gfx1100 Device Vendor Advanced Micro Devices, Inc. Device Vendor ID 0x1002 Device Version OpenCL 2.0 Driver Version 3649.0 (HSA1.1,LC) Device OpenCL C Version OpenCL C 2.0 Device Type GPU Device Board Name (AMD) AMD Radeon RX 7900 XT Device PCI-e ID (AMD) 0x744c Device Topology (AMD) PCI-E, 0000:0b:00.0 Device Profile FULL_PROFILE Device Available Yes Compiler Available Yes Linker Available Yes Max compute units 42 SIMD per compute unit (AMD) 4 SIMD width (AMD) 32 SIMD instruction width (AMD) 1 Max clock frequency 2175MHz Graphics IP (AMD) 11.0 Device Partition (core) Max number of sub-devices 42 Supported partition types None Supported affinity domains (n/a) Max work item dimensions 3 Max work item sizes 1024x1024x1024 Max work group size 256 Preferred work group size (AMD) 256 Max work group size (AMD) 1024 Preferred work group size multiple (kernel) 32 Wavefront width (AMD) 32 Preferred / native vector sizes char 4 / 4 short 2 / 2 int 1 / 1 long 1 / 1 half 1 / 1 (cl_khr_fp16) float 1 / 1 double 1 / 1 (cl_khr_fp64) Half-precision Floating-point support (cl_khr_fp16) Denormals Yes Infinity and NANs Yes Round to nearest Yes Round to zero Yes Round to infinity Yes IEEE754-2008 fused multiply-add Yes Support is emulated in software No Single-precision Floating-point support (core) Denormals Yes Infinity and NANs Yes Round to nearest Yes Round to zero Yes Round to infinity Yes IEEE754-2008 fused multiply-add Yes Support is emulated in software No Correctly-rounded divide and sqrt operations Yes Double-precision Floating-point support (cl_khr_fp64) Denormals Yes Infinity and NANs Yes Round to nearest Yes Round to zero Yes Round to infinity Yes IEEE754-2008 fused multiply-add Yes Support is emulated in software No Address bits 64, Little-Endian Global memory size 21458059264 (19.98GiB) Global free memory (AMD) 20758528 (19.8GiB) 20758528 (19.8GiB) Global memory channels (AMD) 10 Global memory banks per channel (AMD) 4 Global memory bank width (AMD) 256 bytes Error Correction support No Max memory allocation 18239350368 (16.99GiB) Unified memory for Host and Device No Shared Virtual Memory (SVM) capabilities (core) Coarse-grained buffer sharing Yes Fine-grained buffer sharing Yes Fine-grained system sharing No Atomics No Minimum alignment for any data type 128 bytes Alignment of base address 2048 bits (256 bytes) Preferred alignment for atomics SVM 0 bytes Global 0 bytes Local 0 bytes Max size for global variable 18239350368 (16.99GiB) Preferred total size of global vars 21458059264 (19.98GiB) Global Memory cache type Read/Write Global Memory cache size 32768 (32KiB) Global Memory cache line size 128 bytes Image support Yes Max number of samplers per kernel 16 Max size for 1D images from buffer 134217728 pixels Max 1D or 2D image array size 8192 images Base address alignment for 2D image buffers 256 bytes Pitch alignment for 2D image buffers 256 pixels Max 2D image size 16384x16384 pixels Max 3D image size 16384x16384x8192 pixels Max number of read image args 128 Max number of write image args 8 Max number of read/write image args 64 Max number of pipe args 16 Max active pipe reservations 16 Max pipe packet size 1059481184 (1010MiB) Local memory type Local Local memory size 65536 (64KiB) Local memory size per CU (AMD) 65536 (64KiB) Local memory banks (AMD) 32 Max number of constant args 8 Max constant buffer size 18239350368 (16.99GiB) Preferred constant buffer size (AMD) 16384 (16KiB) Max size of kernel argument 1024 Queue properties (on host) Out-of-order execution No Profiling Yes Queue properties (on device) Out-of-order execution Yes Profiling Yes Preferred size 262144 (256KiB) Max size 8388608 (8MiB) Max queues on device 1 Max events on device 1024 Prefer user sync for interop Yes Number of P2P devices (AMD) 0 Profiling timer resolution 1ns Profiling timer offset since Epoch (AMD) 0ns (Thu Jan 1 01:00:00 1970) Execution capabilities Run OpenCL kernels Yes Run native kernels No Thread trace supported (AMD) No Number of async queues (AMD) 8 Max real-time compute queues (AMD) 8 Max real-time compute units (AMD) 42 printf() buffer size 4194304 (4MiB) Built-in kernels (n/a) Device Extensions cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics ... ``
Note that my Darktable build works perfectly fine using ROCM in this machine when run with `RUSTICL_ENABLE=0 darktable`.
Recently released spirv-tools-libs-2025.4-1.fc43 broke rusticl on my system (which is also AMDGPU-based). Downgrading to spirv-tools-libs-2025.3-4.fc43 fixed the issue. Maybe you could try?
Hi @vincent.minder . Indeed the problem is the spirv-tools pkg. Downgraded to the one you suggested and no more crash in either Darktable or clinfo: ``` ➜ ~ clinfo | echo $? 0 ```
Pinging @jfalempe since it looks like a Vulkan SDK issue. There is a new Vulkan SDK version available that might (or not) fix this problem.
Hi Jacobo. Thanks for confirming. Here is more detail about my setup. I encountered the problem after a usual system upgrade on two machines running Fedora/KDE 43: - 1 workstation with AMD Radeon Pro WX 5100 (Polaris 10 - gfx803) - 1 laptop with AMD Ryzen 5 4500U integrated graphics (Vega/Renoir - gfx90c) My custom OpenCL application started segfaulting at launch after the upgrade, with dmesg output revealing the following error (example from laptop, similar on other PC): segfault at 3e ip 00007f7b84c8d280 sp 00007fff63136920 error 4 in libSPIRV-Tools-link.so[7280,7f7b84c86000+1e000] likely on CPU 4 (core 5, socket 0) Downgrading spirv-tools-libs immediately fixed the issue on both systems. Hope this helps. Regards
I can also reproduce this and i can confirm downgrading to spirv-tools-libs-2025.3-4.fc43 fixes the issue. This looks exactly like https://gitlab.freedesktop.org/mesa/mesa/-/issues/14080 According to this, it looks like a rebuild of the mesa package is necessary to fix ABI issues.
Thanks for the details. Yes, that looks like an ABI breakage, we need to rebuild mesa with the new spirv-tools.
See https://bodhi.fedoraproject.org/updates/FEDORA-2026-222c10e880 As mesa is rebuild this is also fixed now.
FEDORA-2026-222c10e880 (DirectX-Headers-1.618.2-1.fc43 and mesa-25.3.4-7.fc43) has been submitted as an update to Fedora 43. https://bodhi.fedoraproject.org/updates/FEDORA-2026-222c10e880