]> git.karo-electronics.de Git - linux-beck.git/commitdiff
IB/hfi1: Avoid large frame size warning
authorLeon Romanovsky <leonro@mellanox.com>
Tue, 31 May 2016 07:54:36 +0000 (10:54 +0300)
committerDoug Ledford <dledford@redhat.com>
Mon, 6 Jun 2016 23:16:52 +0000 (19:16 -0400)
When CONFIG_FRAME_WARN is set to 1024 bytes, which is useful to find
stack consumers, we get a warning in hfi1 driver.

drivers/infiniband/hw/hfi1/affinity.c: In function
‘hfi1_get_proc_affinity’:
drivers/infiniband/hw/hfi1/affinity.c:415:1: warning: the frame size of
1056 bytes is larger than 1024 bytes [-Wframe-larger-than=]

This change removes unneeded buf[1024] declaration and usage.

Fixes: f48ad614c100 ("IB/hfi1: Move driver out of staging")
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Acked-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
drivers/infiniband/hw/hfi1/affinity.c

index 6e7050ab9e16049039406b8cbb207cdfced18109..14d7eeb09be6545f5f21144f0f11ea41c53203a4 100644 (file)
@@ -300,16 +300,15 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node)
        const struct cpumask *node_mask,
                *proc_mask = tsk_cpus_allowed(current);
        struct cpu_mask_set *set = &dd->affinity->proc;
-       char buf[1024];
 
        /*
         * check whether process/context affinity has already
         * been set
         */
        if (cpumask_weight(proc_mask) == 1) {
-               scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(proc_mask));
-               hfi1_cdbg(PROC, "PID %u %s affinity set to CPU %s",
-                         current->pid, current->comm, buf);
+               hfi1_cdbg(PROC, "PID %u %s affinity set to CPU %*pbl",
+                         current->pid, current->comm,
+                         cpumask_pr_args(proc_mask));
                /*
                 * Mark the pre-set CPU as used. This is atomic so we don't
                 * need the lock
@@ -318,9 +317,9 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node)
                cpumask_set_cpu(cpu, &set->used);
                goto done;
        } else if (cpumask_weight(proc_mask) < cpumask_weight(&set->mask)) {
-               scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(proc_mask));
-               hfi1_cdbg(PROC, "PID %u %s affinity set to CPU set(s) %s",
-                         current->pid, current->comm, buf);
+               hfi1_cdbg(PROC, "PID %u %s affinity set to CPU set(s) %*pbl",
+                         current->pid, current->comm,
+                         cpumask_pr_args(proc_mask));
                goto done;
        }
 
@@ -356,8 +355,8 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node)
        cpumask_or(intrs, intrs, (dd->affinity->rcv_intr.gen ?
                                  &dd->affinity->rcv_intr.mask :
                                  &dd->affinity->rcv_intr.used));
-       scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(intrs));
-       hfi1_cdbg(PROC, "CPUs used by interrupts: %s", buf);
+       hfi1_cdbg(PROC, "CPUs used by interrupts: %*pbl",
+                 cpumask_pr_args(intrs));
 
        /*
         * If we don't have a NUMA node requested, preference is towards
@@ -366,18 +365,16 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node)
        if (node == -1)
                node = dd->node;
        node_mask = cpumask_of_node(node);
-       scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(node_mask));
-       hfi1_cdbg(PROC, "device on NUMA %u, CPUs %s", node, buf);
+       hfi1_cdbg(PROC, "device on NUMA %u, CPUs %*pbl", node,
+                 cpumask_pr_args(node_mask));
 
        /* diff will hold all unused cpus */
        cpumask_andnot(diff, &set->mask, &set->used);
-       scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(diff));
-       hfi1_cdbg(PROC, "unused CPUs (all) %s", buf);
+       hfi1_cdbg(PROC, "unused CPUs (all) %*pbl", cpumask_pr_args(diff));
 
        /* get cpumask of available CPUs on preferred NUMA */
        cpumask_and(mask, diff, node_mask);
-       scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(mask));
-       hfi1_cdbg(PROC, "available cpus on NUMA %s", buf);
+       hfi1_cdbg(PROC, "available cpus on NUMA %*pbl", cpumask_pr_args(mask));
 
        /*
         * At first, we don't want to place processes on the same
@@ -395,8 +392,8 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node)
                cpumask_andnot(diff, &set->mask, &set->used);
                cpumask_andnot(mask, diff, node_mask);
        }
-       scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(mask));
-       hfi1_cdbg(PROC, "possible CPUs for process %s", buf);
+       hfi1_cdbg(PROC, "possible CPUs for process %*pbl",
+                 cpumask_pr_args(mask));
 
        cpu = cpumask_first(mask);
        if (cpu >= nr_cpu_ids) /* empty */