Merge pull request #12338 from lrbison/smcuda_wmb

wenduwan · web-flow · commit a55e9b25d67e · 2024-02-15T10:52:08.000-08:00
btl/smcuda: Add atomic_wmb() before sm_fifo_write
diff --git a/opal/mca/btl/smcuda/btl_smcuda_fifo.h b/opal/mca/btl/smcuda/btl_smcuda_fifo.h
@@ -85,6 +85,8 @@ static void add_pending(struct mca_btl_base_endpoint_t *ep, void *data, bool res
 #define MCA_BTL_SMCUDA_FIFO_WRITE(endpoint_peer, my_smp_rank, peer_smp_rank, hdr, resend,         \
                                   retry_pending_sends, rc)                                        \
     do {                                                                                          \
+        /* memory barrier: ensure writes to the hdr have completed */                             \
+        opal_atomic_wmb();                                                                        \
         sm_fifo_t *_fifo = &(mca_btl_smcuda_component.fifo[peer_smp_rank][FIFO_MAP(my_smp_rank)]);\
                                                                                                   \
         if (retry_pending_sends) {                                                                \

Original file line number	Diff line number	Diff line change
`@@ -85,6 +85,8 @@ static void add_pending(struct mca_btl_base_endpoint_t ep, void data, bool res`
`85`	`85`	`#define MCA_BTL_SMCUDA_FIFO_WRITE(endpoint_peer, my_smp_rank, peer_smp_rank, hdr, resend, \`
`86`	`86`	`retry_pending_sends, rc) \`
`87`	`87`	`do { \`
	`88`	`+ /* memory barrier: ensure writes to the hdr have completed */ \`
	`89`	`+ opal_atomic_wmb(); \`
`88`	`90`	`sm_fifo_t *_fifo = &(mca_btl_smcuda_component.fifo[peer_smp_rank][FIFO_MAP(my_smp_rank)]);\`
`89`	`91`	`\`
`90`	`92`	`if (retry_pending_sends) { \`