From: Jubin John <jubin.j...@intel.com>
The name used to create the verbs txreq cache was not qualified with the unit
number. This causes a panic when destroying the cache on a dual HFI systems.
Create a unique name with the unit number with this patch
Reviewed-by: Mike Marciniszyn <mike.marcinis...@intel.com>
Signed-off-by: Jubin John <jubin.j...@intel.com>
Signed-off-by: Ira Weiny <ira.we...@intel.com>
---
drivers/staging/rdma/hfi1/verbs.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/staging/rdma/hfi1/verbs.c
b/drivers/staging/rdma/hfi1/verbs.c
index b2bb8ce97915..514956af5b5d 100644
--- a/drivers/staging/rdma/hfi1/verbs.c
+++ b/drivers/staging/rdma/hfi1/verbs.c
@@ -129,6 +129,9 @@ static void verbs_sdma_complete(
int status,
int drained);
+/* Length of buffer to create verbs txreq cache name */
+#define TXREQ_NAME_LEN 24
+
/*
* Note that it is OK to post send work requests in the SQE and ERR
* states; hfi1_do_send() will process them and generate error
@@ -1899,6 +1902,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
int ret;
size_t lcpysz = IB_DEVICE_NAME_MAX;
u16 descq_cnt;
+ char buf[TXREQ_NAME_LEN];
ret = hfi1_qp_init(dev);
if (ret)
@@ -1952,8 +1956,9 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
descq_cnt = sdma_get_descq_cnt();
+ snprintf(buf, sizeof(buf), "hfi1_%u_vtxreq_cache", dd->unit);
/* SLAB_HWCACHE_ALIGN for AHG */
- dev->verbs_txreq_cache = kmem_cache_create("hfi1_vtxreq_cache",
+ dev->verbs_txreq_cache = kmem_cache_create(buf,
sizeof(struct verbs_txreq),
0, SLAB_HWCACHE_ALIGN,
verbs_txreq_kmem_cache_ctor);
--
1.8.2
_______________________________________________
devel mailing list
de...@linuxdriverproject.org
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
