If hl_mmu_prefetch_cache_range() fails then this code calls
mutex_unlock(&ctx->mmu_lock) when it's no longer holding the mutex.
Fixes: 9e495e24003e ("habanalabs: do MMU prefetch as deferred work")
Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
---
drivers/misc/habanalabs/common/memory.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/misc/habanalabs/common/memory.c b/drivers/misc/habanalabs/common/memory.c
index 663dd7e589d4..d5e6500f8a1f 100644
--- a/drivers/misc/habanalabs/common/memory.c
+++ b/drivers/misc/habanalabs/common/memory.c
@@ -1245,16 +1245,16 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args, u64 *device
rc = map_phys_pg_pack(ctx, ret_vaddr, phys_pg_pack);
if (rc) {
dev_err(hdev->dev, "mapping page pack failed for handle %u\n", handle);
+ mutex_unlock(&ctx->mmu_lock);
goto map_err;
}
rc = hl_mmu_invalidate_cache_range(hdev, false, *vm_type | MMU_OP_SKIP_LOW_CACHE_INV,
ctx->asid, ret_vaddr, phys_pg_pack->total_size);
+ mutex_unlock(&ctx->mmu_lock);
if (rc)
goto map_err;
- mutex_unlock(&ctx->mmu_lock);
-
/*
* prefetch is done upon user's request. it is performed in WQ as and so can
* be outside the MMU lock. the operation itself is already protected by the mmu lock
@@ -1283,8 +1283,6 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args, u64 *device
return rc;
map_err:
- mutex_unlock(&ctx->mmu_lock);
-
if (add_va_block(hdev, va_range, ret_vaddr,
ret_vaddr + phys_pg_pack->total_size - 1))
dev_warn(hdev->dev,
--
2.35.1
