happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

model: assert nextn_predict_layers to prevent underflow (#20783) 

Address GHSA-645x-v54x-34w8.

When nextn_predict_layers >= n_layer, n_layer - nextn_predict_layers
can underflow (unsigned wrap), which corrupts n_layer_kv_from_start.

Assert nextn_predict_layers immediately after parsing the GGUF key.

Found-by: Pwno
This commit is contained in:
Ruikai Peng 2026-03-20 17:17:58 +08:00 committed by GitHub
parent ab9d4c3678
commit 3adbef7776
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/llama-model.cpp
View File

@ -1673,6 +1673,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
// NextN/MTP parameters (GLM-OCR) // NextN/MTP parameters (GLM-OCR)
ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false); ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false);
GGML_ASSERT(hparams.nextn_predict_layers < hparams.n_layer && "nextn_predict_layers must be < n_layer");
// TODO: when MTP is implemented, this should probably be updated if needed // TODO: when MTP is implemented, this should probably be updated if needed
hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers; hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers;
@ -1706,6 +1707,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
// NextN/MTP parameters // NextN/MTP parameters
ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false); ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false);
GGML_ASSERT(hparams.nextn_predict_layers < hparams.n_layer && "nextn_predict_layers must be < n_layer");
// TODO: when MTP is implemented, this should probably be updated if needed // TODO: when MTP is implemented, this should probably be updated if needed
hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers; hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers;
@ -1752,6 +1754,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
// NextN/MTP parameters // NextN/MTP parameters
ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false); ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false);
GGML_ASSERT(hparams.nextn_predict_layers < hparams.n_layer && "nextn_predict_layers must be < n_layer");
// TODO: when MTP is implemented, this should probably be updated if needed // TODO: when MTP is implemented, this should probably be updated if needed
hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers; hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers;
@ -1926,6 +1929,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
ml.get_key(LLM_KV_LEADING_DENSE_BLOCK_COUNT, hparams.n_layer_dense_lead, false); ml.get_key(LLM_KV_LEADING_DENSE_BLOCK_COUNT, hparams.n_layer_dense_lead, false);
ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false); ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false);
GGML_ASSERT(hparams.nextn_predict_layers < hparams.n_layer && "nextn_predict_layers must be < n_layer");
switch (hparams.n_layer) { switch (hparams.n_layer) {
case 32: type = LLM_TYPE_30B_A3B; break; case 32: type = LLM_TYPE_30B_A3B; break;
@ -2108,6 +2112,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
ml.get_key(LLM_KV_EXPERT_WEIGHTS_NORM, hparams.expert_weights_norm, false); ml.get_key(LLM_KV_EXPERT_WEIGHTS_NORM, hparams.expert_weights_norm, false);
ml.get_key(LLM_KV_EXPERT_GATING_FUNC, hparams.expert_gating_func); ml.get_key(LLM_KV_EXPERT_GATING_FUNC, hparams.expert_gating_func);
ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false); ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.nextn_predict_layers, false);
GGML_ASSERT(hparams.nextn_predict_layers < hparams.n_layer && "nextn_predict_layers must be < n_layer");
// TODO: when MTP is implemented, this should probably be updated if needed // TODO: when MTP is implemented, this should probably be updated if needed
hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers; hparams.n_layer_kv_from_start = hparams.n_layer - hparams.nextn_predict_layers;