selftests: net: bump default cmd() timeout to 20 seconds
Subsystems (3)
Jakub Kicinski
maintainerPATCH 1/1 selftests: net: bump default cmd() timeout to 20 seconds
We always used 5 sec as the default command timeout. But soon after it was introduced, David effectively made us ignore the timeout (it was passed to process.communicate() as the wrong argument). Gal recently fixed that, but turns out the 5 sec is not enough for a lot of tests and setups. The fix regressed regressions.
In particular running reconfig commands (e.g. XDP attach) on mlx5 with 32 rings and 9k MTU, on a heavily-debug-enabled kernel takes more than 5 sec. The XDP installation command will time out after 5 sec but since the sleeps in the kernel are non interruptible the command finishes anyway, leaving the XDP program attached, but with non-zero exit code. defer()ed cleanups are not installed, breaking the environment for subsequent tests.
Since "install XDP" is a pretty normal command a "point fix" does not seem appropriate. 32 rings is a fairly reasonable config, too, so we should just increase the timeout to 20 sec.
There's no real reason behind the value of 20.
---
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
--- tools/testing/selftests/net/lib/py/utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tools/testing/selftests/net/lib/py/utils.py +2 -2
@@ -44,7 +44,7 @@ import time Use bkg() instead to run a command in the background. """ def __init__(self, comm, shell=None, fail=True, expect_fail=False, ns=None, - background=False, host=None, timeout=5, ksft_ready=None, + background=False, host=None, timeout=20, ksft_ready=None, ksft_wait=None): if ns: if hasattr(ns, 'user_ns_path'):
@@ -113,7 +113,7 @@ import time return stdout, stderr - def process(self, terminate=True, fail=None, expect_fail=False, timeout=5): + def process(self, terminate=True, fail=None, expect_fail=False, timeout=20): if fail is None: fail = not terminate
Signature · 1 line
2.54.0
Quoted 32 lines · Jakub Kicinski
We always used 5 sec as the default command timeout. But soon after it was introduced, David effectively made us ignore the timeout (it was passed to
process.communicate()as the wrong argument). Gal recently fixed that, but turns out the 5 sec is not enough for a lot of tests and setups. The fix regressed regressions.In particular running reconfig commands (e.g. XDP attach) on mlx5 with 32 rings and 9k MTU, on a heavily-debug-enabled kernel takes more than 5 sec. The XDP installation command will time out after 5 sec but since the sleeps in the kernel are non interruptible the command finishes anyway, leaving the XDP program attached, but with non-zero exit code. defer()ed cleanups are not installed, breaking the environment for subsequent tests.
Since "install XDP" is a pretty normal command a "point fix" does not seem appropriate. 32 rings is a fairly reasonable config, too, so we should just increase the timeout to 20 sec.
There's no real reason behind the value of 20.
Fixes 1cf270424218 ("net: selftest: add test for netdev netlink queue-get API")Fixes f0bd19316663 ("selftests: net: fix timeout passed as positional argument to communicate()")Signed-off-by Jakub Kicinski <[email protected]>---
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
Quoted 4 lines
--- tools/testing/selftests/net/lib/py/utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Quoted 26 lines
tools/testing/selftests/net/lib/py/utils.py +2
-3@@ -44,7 +44,7 @@ import time Use bkg() instead to run a command in the background. """ def __init__(self, comm, shell=None, fail=True, expect_fail=False, ns=None, - background=False, host=None, timeout=5, ksft_ready=None, + background=False, host=None, timeout=20, ksft_ready=None, ksft_wait=None): if ns: if hasattr(ns, 'user_ns_path'):@@ -113,7 +113,7 @@ import time return stdout, stderr - def process(self, terminate=True, fail=None, expect_fail=False, timeout=5): + def process(self, terminate=True, fail=None, expect_fail=False, timeout=20): if fail is None: fail = not terminate --2.54.0
Quoted 18 lines · Jakub Kicinski
We always used 5 sec as the default command timeout. But soon after it was introduced, David effectively made us ignore the timeout (it was passed to
process.communicate()as the wrong argument). Gal recently fixed that, but turns out the 5 sec is not enough for a lot of tests and setups. The fix regressed regressions.In particular running reconfig commands (e.g. XDP attach) on mlx5 with 32 rings and 9k MTU, on a heavily-debug-enabled kernel takes more than 5 sec. The XDP installation command will time out after 5 sec but since the sleeps in the kernel are non interruptible the command finishes anyway, leaving the XDP program attached, but with non-zero exit code. defer()ed cleanups are not installed, breaking the environment for subsequent tests.
Since "install XDP" is a pretty normal command a "point fix" does not seem appropriate. 32 rings is a fairly reasonable config, too, so we should just increase the timeout to 20 sec.
Quoted 5 lines
There's no real reason behind the value of 20.
Fixes 1cf270424218 ("net: selftest: add test for netdev netlink queue-get API")Fixes f0bd19316663 ("selftests: net: fix timeout passed as positional argument to communicate()")Signed-off-by Jakub Kicinski <[email protected]>
Quoted 31 lines · Jakub Kicinski
We always used 5 sec as the default command timeout. But soon after it was introduced, David effectively made us ignore the timeout (it was passed to
process.communicate()as the wrong argument). Gal recently fixed that, but turns out the 5 sec is not enough for a lot of tests and setups. The fix regressed regressions.In particular running reconfig commands (e.g. XDP attach) on mlx5 with 32 rings and 9k MTU, on a heavily-debug-enabled kernel takes more than 5 sec. The XDP installation command will time out after 5 sec but since the sleeps in the kernel are non interruptible the command finishes anyway, leaving the XDP program attached, but with non-zero exit code. defer()ed cleanups are not installed, breaking the environment for subsequent tests.
Since "install XDP" is a pretty normal command a "point fix" does not seem appropriate. 32 rings is a fairly reasonable config, too, so we should just increase the timeout to 20 sec.
There's no real reason behind the value of 20.
Fixes 1cf270424218 ("net: selftest: add test for netdev netlink queue-get API")Fixes f0bd19316663 ("selftests: net: fix timeout passed as positional argument to communicate()")Signed-off-by Jakub Kicinski <[email protected]>---
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
Quoted 4 lines
--- tools/testing/selftests/net/lib/py/utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Quoted 22 lines
tools/testing/selftests/net/lib/py/utils.py +2
-2@@ -44,7 +44,7 @@ import time Use bkg() instead to run a command in the background. """ def __init__(self, comm, shell=None, fail=True, expect_fail=False, ns=None, - background=False, host=None, timeout=5, ksft_ready=None, + background=False, host=None, timeout=20, ksft_ready=None, ksft_wait=None): if ns: if hasattr(ns, 'user_ns_path'):@@ -113,7 +113,7 @@ import time return stdout, stderr - def process(self, terminate=True, fail=None, expect_fail=False, timeout=5): + def process(self, terminate=True, fail=None, expect_fail=False, timeout=20): if fail is None: fail = not terminate
This patch was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:
Hello:
This patch was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:
Quoted 14 lines · On Mon, 29 Jun 2026 16:33:48 -0700 you wrote:
We always used 5 sec as the default command timeout. But soon after it was introduced, David effectively made us ignore the timeout (it was passed to
process.communicate()as the wrong argument). Gal recently fixed that, but turns out the 5 sec is not enough for a lot of tests and setups. The fix regressed regressions.In particular running reconfig commands (e.g. XDP attach) on mlx5 with 32 rings and 9k MTU, on a heavily-debug-enabled kernel takes more than 5 sec. The XDP installation command will time out after 5 sec but since the sleeps in the kernel are non interruptible the command finishes anyway, leaving the XDP program attached, but with non-zero exit code. defer()ed cleanups are not installed, breaking the environment for subsequent tests.
Quoted 1 line
[...]
Here is the summary with links:
- [net] selftests: net: bump default cmd() timeout to 20 seconds
https://git.kernel.org/netdev/net/c/57bb59ab6fa3
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
Quoted 6 lines · Jakub Kicinski
We always used 5 sec as the default command timeout. But soon after it was introduced, David effectively made us ignore the timeout (it was passed to
process.communicate()as the wrong argument). Gal recently fixed that, but turns out the 5 sec is not enough for a lot of tests and setups. The fix regressed regressions.
Quoted 8 lines
In particular running reconfig commands (e.g. XDP attach) on mlx5 with 32 rings and 9k MTU, on a heavily-debug-enabled kernel takes more than 5 sec. The XDP installation command will time out after 5 sec but since the sleeps in the kernel are non interruptible the command finishes anyway, leaving the XDP program attached, but with non-zero exit code. defer()ed cleanups are not installed, breaking the environment for subsequent tests.
Hi Jakub,
We've improved the performance of configuration change operations over the past 2 years. We have more patches to be submitted soon, in addition to planned ones.
I'd be glad if you could share some details about the NIC and FW version for which you hit this 5sec timeout.
Patch LGTM.
Thanks.
Quoted 43 lines
Since "install XDP" is a pretty normal command a "point fix" does not seem appropriate. 32 rings is a fairly reasonable config, too, so we should just increase the timeout to 20 sec.
There's no real reason behind the value of 20.
Fixes 1cf270424218 ("net: selftest: add test for netdev netlink queue-get API")Fixes f0bd19316663 ("selftests: net: fix timeout passed as positional argument to communicate()")Signed-off-by Jakub Kicinski <[email protected]>---
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]
CC: [email protected]--- tools/testing/selftests/net/lib/py/utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)tools/testing/selftests/net/lib/py/utils.py +2
-2@@ -44,7 +44,7 @@ import time Use bkg() instead to run a command in the background. """ def __init__(self, comm, shell=None, fail=True, expect_fail=False, ns=None, - background=False, host=None, timeout=5, ksft_ready=None, + background=False, host=None, timeout=20, ksft_ready=None, ksft_wait=None): if ns: if hasattr(ns, 'user_ns_path'):@@ -113,7 +113,7 @@ import time return stdout, stderr - def process(self, terminate=True, fail=None, expect_fail=False, timeout=5): + def process(self, terminate=True, fail=None, expect_fail=False, timeout=20): if fail is None: fail = not terminate