github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 08:02:14 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #2066] Switching from CUDA to CPU runner causes segmentation fault #26956

New Issue
Closed
opened 2026-04-22 03:45:11 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-22 03:45:11 -05:00
Owner
Copy Link

Originally created by @jmorganca on GitHub (Jan 19, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/2066

Originally assigned to: @dhiltgen on GitHub.

This is only currently an issue on main

2024/01/19 04:46:40 routes.go:76: INFO changing loaded model
2024/01/19 04:46:40 gpu.go:136: INFO CUDA Compute Capability detected: 8.9
2024/01/19 04:46:40 gpu.go:136: INFO CUDA Compute Capability detected: 8.9
2024/01/19 04:46:40 cpu_common.go:11: INFO CPU has AVX2
loading library /tmp/ollama2500718665/cpu_avx2/libext_server.so
2024/01/19 04:46:40 dyn_ext_server.go:90: INFO Loading Dynamic llm server: /tmp/ollama2500718665/cpu_avx2/libext_server.so
2024/01/19 04:46:40 dyn_ext_server.go:139: INFO Initializing llama server
SIGSEGV: segmentation violation
PC=0x7f811abadac8 m=5 sigcode=1
signal arrived during cgo execution

goroutine 14 [syscall]:
runtime.cgocall(0x9b4550, 0xc000a4e808)
	/usr/local/go/src/runtime/cgocall.go:157 +0x4b fp=0xc000a4e7e0 sp=0xc000a4e7a8 pc=0x409b0b
github.com/jmorganca/ollama/llm._Cfunc_dyn_llama_server_init({0x7f80bc000f60, 0x7f805a501b80, 0x7f805a4f3a80, 0x7f805a4f7960, 0x7f805a505650, 0x7f805a4ffba0, 0x7f805a4f7930, 0x7f805a4f3b00, 0x7f805a505e00, 0x7f805a505200, ...}, ...)
	_cgo_gotypes.go:280 +0x45 fp=0xc000a4e808 sp=0xc000a4e7e0 pc=0x7c2a45
github.com/jmorganca/ollama/llm.newDynExtServer.func7(0xae6f80?, 0x6e?)
	/go/src/github.com/jmorganca/ollama/llm/dyn_ext_server.go:142 +0xef fp=0xc000a4e8f8 sp=0xc000a4e808 pc=0x7c3eaf
github.com/jmorganca/ollama/llm.newDynExtServer({0xc000134090, 0x2f}, {0xc0009f4150, _}, {_, _, _}, {0x0, 0x0, 0x0}, ...)
	/go/src/github.com/jmorganca/ollama/llm/dyn_ext_server.go:142 +0xa32 fp=0xc000a4eb88 sp=0xc000a4e8f8 pc=0x7c3bf2
github.com/jmorganca/ollama/llm.newLlmServer({{_, _, _}, {_, _}, {_, _}}, {_, _}, {0x0, ...}, ...)
	/go/src/github.com/jmorganca/ollama/llm/llm.go:147 +0x36a fp=0xc000a4ed48 sp=0xc000a4eb88 pc=0x7c04ea
github.com/jmorganca/ollama/llm.New({0x419c8f?, 0x1000100000100?}, {0xc0009f4150, _}, {_, _, _}, {0x0, 0x0, 0x0}, ...)
	/go/src/github.com/jmorganca/ollama/llm/llm.go:122 +0x6f9 fp=0xc000a4efb8 sp=0xc000a4ed48 pc=0x7bff19
github.com/jmorganca/ollama/server.load(0xc00017e900?, 0xc00017e900, {{0x0, 0x800, 0x200, 0x1, 0x0, 0x0, 0x0, 0x1, ...}, ...}, ...)
	/go/src/github.com/jmorganca/ollama/server/routes.go:83 +0x3a5 fp=0xc000a4f138 sp=0xc000a4efb8 pc=0x9908a5
github.com/jmorganca/ollama/server.ChatHandler(0xc00007c100)
	/go/src/github.com/jmorganca/ollama/server/routes.go:1071 +0x828 fp=0xc000a4f748 sp=0xc000a4f138 pc=0x99b1e8
github.com/gin-gonic/gin.(*Context).Next(...)
Originally created by @jmorganca on GitHub (Jan 19, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/2066 Originally assigned to: @dhiltgen on GitHub. This is only currently an issue on `main` ``` 2024/01/19 04:46:40 routes.go:76: INFO changing loaded model 2024/01/19 04:46:40 gpu.go:136: INFO CUDA Compute Capability detected: 8.9 2024/01/19 04:46:40 gpu.go:136: INFO CUDA Compute Capability detected: 8.9 2024/01/19 04:46:40 cpu_common.go:11: INFO CPU has AVX2 loading library /tmp/ollama2500718665/cpu_avx2/libext_server.so 2024/01/19 04:46:40 dyn_ext_server.go:90: INFO Loading Dynamic llm server: /tmp/ollama2500718665/cpu_avx2/libext_server.so 2024/01/19 04:46:40 dyn_ext_server.go:139: INFO Initializing llama server SIGSEGV: segmentation violation PC=0x7f811abadac8 m=5 sigcode=1 signal arrived during cgo execution goroutine 14 [syscall]: runtime.cgocall(0x9b4550, 0xc000a4e808) /usr/local/go/src/runtime/cgocall.go:157 +0x4b fp=0xc000a4e7e0 sp=0xc000a4e7a8 pc=0x409b0b github.com/jmorganca/ollama/llm._Cfunc_dyn_llama_server_init({0x7f80bc000f60, 0x7f805a501b80, 0x7f805a4f3a80, 0x7f805a4f7960, 0x7f805a505650, 0x7f805a4ffba0, 0x7f805a4f7930, 0x7f805a4f3b00, 0x7f805a505e00, 0x7f805a505200, ...}, ...) _cgo_gotypes.go:280 +0x45 fp=0xc000a4e808 sp=0xc000a4e7e0 pc=0x7c2a45 github.com/jmorganca/ollama/llm.newDynExtServer.func7(0xae6f80?, 0x6e?) /go/src/github.com/jmorganca/ollama/llm/dyn_ext_server.go:142 +0xef fp=0xc000a4e8f8 sp=0xc000a4e808 pc=0x7c3eaf github.com/jmorganca/ollama/llm.newDynExtServer({0xc000134090, 0x2f}, {0xc0009f4150, _}, {_, _, _}, {0x0, 0x0, 0x0}, ...) /go/src/github.com/jmorganca/ollama/llm/dyn_ext_server.go:142 +0xa32 fp=0xc000a4eb88 sp=0xc000a4e8f8 pc=0x7c3bf2 github.com/jmorganca/ollama/llm.newLlmServer({{_, _, _}, {_, _}, {_, _}}, {_, _}, {0x0, ...}, ...) /go/src/github.com/jmorganca/ollama/llm/llm.go:147 +0x36a fp=0xc000a4ed48 sp=0xc000a4eb88 pc=0x7c04ea github.com/jmorganca/ollama/llm.New({0x419c8f?, 0x1000100000100?}, {0xc0009f4150, _}, {_, _, _}, {0x0, 0x0, 0x0}, ...) /go/src/github.com/jmorganca/ollama/llm/llm.go:122 +0x6f9 fp=0xc000a4efb8 sp=0xc000a4ed48 pc=0x7bff19 github.com/jmorganca/ollama/server.load(0xc00017e900?, 0xc00017e900, {{0x0, 0x800, 0x200, 0x1, 0x0, 0x0, 0x0, 0x1, ...}, ...}, ...) /go/src/github.com/jmorganca/ollama/server/routes.go:83 +0x3a5 fp=0xc000a4f138 sp=0xc000a4efb8 pc=0x9908a5 github.com/jmorganca/ollama/server.ChatHandler(0xc00007c100) /go/src/github.com/jmorganca/ollama/server/routes.go:1071 +0x828 fp=0xc000a4f748 sp=0xc000a4f138 pc=0x99b1e8 github.com/gin-gonic/gin.(*Context).Next(...) ```
GiteaMirror added the bug label 2026-04-22 03:45:11 -05:00
GiteaMirror commented 2026-04-22 03:45:13 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jan 19, 2024):

Repro scenario: On a 4G card.

# Get the GPU runner loaded
% ollama run phi hello
...
% curl http://localhost:11434/api/generate -d '{
  "model": "mistral",
  "prompt": "hello",
  "stream": false, "options": {"num_ctx": 65536}
}'

There's some piece of global state that's lingering and not getting cleaned up. Possibly related to #1848

<!-- gh-comment-id:1900844798 --> @dhiltgen commented on GitHub (Jan 19, 2024): Repro scenario: On a 4G card. ``` # Get the GPU runner loaded % ollama run phi hello ... % curl http://localhost:11434/api/generate -d '{ "model": "mistral", "prompt": "hello", "stream": false, "options": {"num_ctx": 65536} }' ``` There's some piece of global state that's lingering and not getting cleaned up. Possibly related to #1848
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#26956
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?