X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/095e176632bbf81d28a239742a1ecce12404bd2d..e672f484160faae900fb7f7e281d06952fd35d28:/lib/crunchrun/cuda.go

diff --git a/lib/crunchrun/cuda.go b/lib/crunchrun/cuda.go
index c693dbcb96..f91a5c62cd 100644
--- a/lib/crunchrun/cuda.go
+++ b/lib/crunchrun/cuda.go
@@ -5,13 +5,15 @@
 package crunchrun
 
 import (
+	"fmt"
+	"io"
 	"os/exec"
 )
 
 // nvidiaModprobe makes sure all the nvidia kernel modules and devices
 // are set up.  If we don't have all the modules/devices set up we get
 // "CUDA_ERROR_UNKNOWN".
-func nvidiaModprobe(writer *ThrottledLogger) {
+func nvidiaModprobe(writer io.Writer) {
 	// The underlying problem is that when normally running
 	// directly on the host, the CUDA SDK will automatically
 	// detect and set up the devices on demand.  However, when
@@ -42,7 +44,7 @@ func nvidiaModprobe(writer *ThrottledLogger) {
 	nvidiaSmi.Stderr = writer
 	err := nvidiaSmi.Run()
 	if err != nil {
-		writer.Printf("Warning %v: %v", nvidiaSmi.Args, err)
+		fmt.Fprintf(writer, "Warning %v: %v\n", nvidiaSmi.Args, err)
 	}
 
 	// Load the kernel modules & devices associated with
@@ -63,7 +65,7 @@ func nvidiaModprobe(writer *ThrottledLogger) {
 		nvmodprobe.Stderr = writer
 		err = nvmodprobe.Run()
 		if err != nil {
-			writer.Printf("Warning %v: %v", nvmodprobe.Args, err)
+			fmt.Fprintf(writer, "Warning %v: %v\n", nvmodprobe.Args, err)
 		}
 	}
 }