Skip to contentSkip to conversation

selftests/uevent: increase __UEVENT_BUFFER_SIZE to avoid ENOBUFS on busy systems

PATCH Applied Guixin Liu 2 participants 3 messages Jun 8 – Jun 10 Kselftest

lore
Share & tools
lore

Guixin Liu

author
#0 lore raw

PATCH 1/1 selftests/uevent: increase __UEVENT_BUFFER_SIZE to avoid ENOBUFS on busy systems

The kselftests case uevent.uevent_filtering fails reproducibly on busy systems (e.g. Intel EMR / AMD servers) with:

No buffer space available - Failed to receive uevent

The listener binds the NETLINK_KOBJECT_UEVENT socket to all 32 multicast groups (nl_groups = -1) but only sets SO_RCVBUF to 4 KiB (__UEVENT_BUFFER_SIZE = 2048 * 2). On hosts with many devices, the kernel and userspace daemons (udev/systemd) constantly emit uevents on multiple groups, plus the test itself triggers 10 add events in a row. The 4 KiB receive buffer overflows before the listener can drain it, recvmsg() returns -ENOBUFS, and the test bails out as failure.

Increase __UEVENT_BUFFER_SIZE to 1 MiB so the receive buffer is large enough to absorb the burst of uevents on busy systems. After this change the test passes consistently across dozens of runs on Intel EMR and AMD platforms.

---
 tools/testing/selftests/uevent/uevent_filtering.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/testing/selftests/uevent/uevent_filtering.c +1 -1
index 974b076f9235..33a09f66d7e2 100644
@@ -22,7 +22,7 @@
 #include "kselftest_harness.h"
 
 #define __DEV_FULL "/sys/devices/virtual/mem/full/uevent"
-#define __UEVENT_BUFFER_SIZE (2048 * 2)
+#define __UEVENT_BUFFER_SIZE (1024 * 1024)
 #define __UEVENT_HEADER "add@/devices/virtual/mem/full"
 #define __UEVENT_HEADER_LEN sizeof("add@/devices/virtual/mem/full")
 #define __UEVENT_LISTEN_ALL -1
Signature · 1 line
2.43.7
#1 lore raw
Quoted 13 lines · On Mon, 8 Jun 2026 13:39:34 +0800 Guixin Liu <[email protected]> wrote:

The kselftests case uevent.uevent_filtering fails reproducibly on busy systems (e.g. Intel EMR / AMD servers) with:

No buffer space available - Failed to receive uevent

The listener binds the NETLINK_KOBJECT_UEVENT socket to all 32 multicast groups (nl_groups = -1) but only sets SO_RCVBUF to 4 KiB (__UEVENT_BUFFER_SIZE = 2048 * 2). On hosts with many devices, the kernel and userspace daemons (udev/systemd) constantly emit uevents on multiple groups, plus the test itself triggers 10 add events in a row. The 4 KiB receive buffer overflows before the listener can drain it, recvmsg() returns -ENOBUFS, and the test bails out as failure.

Quoted 5 lines

Increase __UEVENT_BUFFER_SIZE to 1 MiB so the receive buffer is large enough to absorb the burst of uevents on busy systems. After this change the test passes consistently across dozens of runs on Intel EMR and AMD platforms.

Fair enough.

tools/testing/selftests/uevent/uevent_filtering.c line 28

Quoted 11 lines
tools/testing/selftests/uevent/uevent_filtering.c +1 -1
@@ -22,7 +22,7 @@
 #include "kselftest_harness.h"
 
 #define __DEV_FULL "/sys/devices/virtual/mem/full/uevent"
-#define __UEVENT_BUFFER_SIZE (2048 * 2)
+#define __UEVENT_BUFFER_SIZE (1024 * 1024)
 #define __UEVENT_HEADER "add@/devices/virtual/mem/full"
 #define __UEVENT_HEADER_LEN sizeof("add@/devices/virtual/mem/full")
 #define __UEVENT_LISTEN_ALL -1

scripts/get_maintainer.sh isn't very useful with selftests code. But git-blame can be used to find out who to Cc (hi, Christian!).

author
#2 lore raw
Quoted 27 lines · Andrew Morton
Quoted 13 lines · On Mon, 8 Jun 2026 13:39:34 +0800 Guixin Liu <[email protected]> wrote:

The kselftests case uevent.uevent_filtering fails reproducibly on busy systems (e.g. Intel EMR / AMD servers) with:

No buffer space available - Failed to receive uevent

The listener binds the NETLINK_KOBJECT_UEVENT socket to all 32 multicast groups (nl_groups = -1) but only sets SO_RCVBUF to 4 KiB (__UEVENT_BUFFER_SIZE = 2048 * 2). On hosts with many devices, the kernel and userspace daemons (udev/systemd) constantly emit uevents on multiple groups, plus the test itself triggers 10 add events in a row. The 4 KiB receive buffer overflows before the listener can drain it, recvmsg() returns -ENOBUFS, and the test bails out as failure.

Quoted 5 lines

Increase __UEVENT_BUFFER_SIZE to 1 MiB so the receive buffer is large enough to absorb the burst of uevents on busy systems. After this change the test passes consistently across dozens of runs on Intel EMR and AMD platforms.

Fair enough.

Quoted 5 lines
tools/testing/selftests/uevent/uevent_filtering.c +0 -0
@@ -22,7 +22,7 @@
  #include "kselftest_harness.h"
 
Quoted 8 lines

tools/testing/selftests/uevent/uevent_filtering.c line 28

Quoted 6 lines
tools/testing/selftests/uevent/uevent_filtering.c +1 -1
  #define __DEV_FULL "/sys/devices/virtual/mem/full/uevent"
-#define __UEVENT_BUFFER_SIZE (2048 * 2)
+#define __UEVENT_BUFFER_SIZE (1024 * 1024)
  #define __UEVENT_HEADER "add@/devices/virtual/mem/full"
  #define __UEVENT_HEADER_LEN sizeof("add@/devices/virtual/mem/full")
  #define __UEVENT_LISTEN_ALL -1

scripts/get_maintainer.sh isn't very useful with selftests code. But git-blame can be used to find out who to Cc (hi, Christian!).

Got it, thanks.

In addtion, It's nice to get a rb tag.

Best Regards,
Guixin Liu