tests/drv_module_reload_basic: Don't use rmmod exit code when reloading the module.
PATCH 1/1 tests/drv_module_reload_basic: Don't use rmmod exit code when reloading the module.
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.
--- tests/drv_module_reload_basic | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tests/drv_module_reload_basic +2 -2
@@ -30,7 +30,7 @@ function reload() { #ignore errors in ips - gen5 only rmmod intel_ips &> /dev/null - rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915 #ignore errors in intel-gtt, often built-in rmmod intel-gtt &> /dev/null # drm may be used by other devices (nouveau, radeon, udl, etc)
@@ -76,7 +76,7 @@ finish_load || exit $? # Repeat the module reload trying to to generate faults for i in $(seq 1 4); do - reload inject_load_failure=$i + reload inject_load_failure=$i || exit $? done reload || exit $?
Signature · 6 lines
2.5.0 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 21 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/null
tests/drv_module_reload_basic function reload() { · line 33
Quoted 2 lines
tests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915
Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
tests/drv_module_reload_basic finish_load || exit $? · line 79
Quoted 10 lines
#ignore errors in intel-gtt, often built-in rmmod intel-gtt &>
/dev/null# drm may be used by other devices (nouveau, radeon, udl, etc)tests/drv_module_reload_basic +0
-0@@ -76,7 +76,7 @@ finish_load || exit $?
# Repeat the module reload trying to to generate faults for i in $(seq 1 4); dotests/drv_module_reload_basic +1
-1- reload inject_load_failure=$i + reload inject_load_failure=$i || exit $?
The idea was to keep the system in a working state even in case of failure here, so I'd still attempt a normal reload before exiting with failure.
--Imre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 13 lines · Imre Deak
Quoted 12 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)
Quoted 14 lines
tests/drv_module_reload_basic function reload() { · line 33
Quoted 11 lines
tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
If we can't unload, we can perform the modprobe testing. The system is not in a state suitable for testing so skip or fail. If we are certain that the rmmod failure is a bug, fail, if it merely something like the system doesn't support module unloading, skip.
Continuing on after failure to unload is not a good idea. -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 39 lines · Chris Wilson
Quoted 14 lines · Imre Deak
Quoted 13 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless.
Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)Quoted 15 lines
tests/drv_module_reload_basic function reload() { · line 33
Quoted 11 lines
tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like
the correct thing to do.If we can't unload, we can perform the modprobe testing. The system is
not in a state suitable for testing so skip or fail. If we are certain
that the rmmod failure is a bug, fail, if it merely something like the
system doesn't support module unloading, skip.
Quoted 1 line
Continuing on after failure to unload is not a good idea.
I meant continuing here and depending on the lsmod check later to exit.
Quoted 2 lines
-Chris
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 29 lines · Chris Wilson
Quoted 13 lines · Imre Deak
Quoted 12 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)Quoted 14 lines
tests/drv_module_reload_basic function reload() { · line 33
Quoted 11 lines
tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
Quoted 4 lines
If we can't unload, we can perform the modprobe testing. The system is not in a state suitable for testing so skip or fail. If we are certain that the rmmod failure is a bug, fail, if it merely something like the system doesn't support module unloading, skip.
I've seen this couple of times in the CI...and went mostly mostly unnoticed, wanted to make it hard-fail so it easy to spot when it happens again. Although it might be cases where the module is built-in this is not the case in CI.
Quoted 6 lines
Continuing on after failure to unload is not a good idea. -Chris
Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 30 lines · Marius Vlad
Quoted 29 lines · Chris Wilson
Quoted 27 lines · Imre Deak
Quoted 21 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic function reload() { · line 33
Quoted 2 lines
tests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
Quoted 8 lines
Quoted 4 lines
If we can't unload, we can perform the modprobe testing. The system is not in a state suitable for testing so skip or fail. If we are certain that the rmmod failure is a bug, fail, if it merely something like the system doesn't support module unloading, skip.
I've seen this couple of times in the CI...and went mostly mostly unnoticed, wanted to make it hard-fail so it easy to spot when it happens again. Although it might be cases where the module is built-in this is not the case in CI.
The || SKIP was added in
commit f14d56c42d9e43df2790465aba6a2ea2593418fc
Author: Chris Wilson <[email protected]> Date: Fri Mar 11 21:25:48 2016 +0000
igt/drv_module_reload_basic: Rinse and repeat with addition module parameters
and apparently not intentionally. Please cite that commit using Fixes: and ack on your patch from my side.
-Daniel
Quoted 7 lines
Quoted 7 lines
Continuing on after failure to unload is not a good idea. -Chris
Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 42 lines · Daniel Vetter
Quoted 30 lines · Marius Vlad
Quoted 29 lines · Chris Wilson
Quoted 27 lines · Imre Deak
Quoted 21 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic function reload() { · line 33
Quoted 2 lines
tests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
Quoted 8 lines
Quoted 4 lines
If we can't unload, we can perform the modprobe testing. The system is not in a state suitable for testing so skip or fail. If we are certain that the rmmod failure is a bug, fail, if it merely something like the system doesn't support module unloading, skip.
I've seen this couple of times in the CI...and went mostly mostly unnoticed, wanted to make it hard-fail so it easy to spot when it happens again. Although it might be cases where the module is built-in this is not the case in CI.
The || SKIP was added in
Quoted 7 lines
commit
f14d56c42d9e43df2790465aba6a2ea2593418fcAuthor: Chris Wilson <[email protected]> Date: Fri Mar 11 21:25:48 2016 +0000igt/drv_module_reload_basic: Rinse and repeat with addition module parameters
and apparently not intentionally.
It was, because the test wasn't behaving properly on my setup where I had disabled module unloading. If we can't unload the module, we have to SKIP the test.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 47 lines · Chris Wilson
Quoted 46 lines · Daniel Vetter
Quoted 30 lines · Marius Vlad
Quoted 29 lines · Chris Wilson
Quoted 27 lines · Imre Deak
Quoted 21 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic function reload() { · line 33
Quoted 2 lines
tests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
Quoted 8 lines
Quoted 4 lines
If we can't unload, we can perform the modprobe testing. The system is not in a state suitable for testing so skip or fail. If we are certain that the rmmod failure is a bug, fail, if it merely something like the system doesn't support module unloading, skip.
I've seen this couple of times in the CI...and went mostly mostly unnoticed, wanted to make it hard-fail so it easy to spot when it happens again. Although it might be cases where the module is built-in this is not the case in CI.
The || SKIP was added in
commit
f14d56c42d9e43df2790465aba6a2ea2593418fcAuthor: Chris Wilson <[email protected]> Date: Fri Mar 11 21:25:48 2016 +0000
Quoted 7 lines
Quoted 3 lines
igt/drv_module_reload_basic: Rinse and repeat with addition module parameters
and apparently not intentionally.
It was, because the test wasn't behaving properly on my setup where I had disabled module unloading. If we can't unload the module, we have to SKIP the test.
The problem is that we actually have the test split up incorrectly. Instead of doing a
preparation phase;
for_each_subtest
load; check; unload check;
restoration
we have
for_each_subtest;
unload
load
check
and so we are doingthe tail of the first test in the second. -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 47 lines · Chris Wilson
Quoted 46 lines · Daniel Vetter
Quoted 30 lines · Marius Vlad
Quoted 29 lines · Chris Wilson
Quoted 27 lines · Imre Deak
Quoted 21 lines · Marius Vlad
Either we return $IGT_EXIT_FAILURE or remove it entirely (like in this
patch). If rmmod returns non-zero (i.e., Module: i915 is still in use), reload
will bail with $IGT_EXIT_SKIP, making the check with lsmod useless. Also use the return value in the fault-injection loop.Signed-off-by Marius Vlad <[email protected]>---
tests/drv_module_reload_basic | 4 ++--1 file changed, 2 insertions(+), 2 deletions(-)tests/drv_module_reload_basic +0
-0@@ -30,7 +30,7 @@ function reload() {
#ignore errors in ips - gen5 only
rmmodintel_ips&>/dev/nulltests/drv_module_reload_basic function reload() { · line 33
Quoted 2 lines
tests/drv_module_reload_basic +1
-1- rmmod i915 || return $IGT_EXIT_SKIP + rmmod i915Not sure what was the reason to bail out here, continuing seems like the correct thing to do.
Quoted 8 lines
Quoted 4 lines
If we can't unload, we can perform the modprobe testing. The system is not in a state suitable for testing so skip or fail. If we are certain that the rmmod failure is a bug, fail, if it merely something like the system doesn't support module unloading, skip.
I've seen this couple of times in the CI...and went mostly mostly unnoticed, wanted to make it hard-fail so it easy to spot when it happens again. Although it might be cases where the module is built-in this is not the case in CI.
The || SKIP was added in
commit
f14d56c42d9e43df2790465aba6a2ea2593418fcAuthor: Chris Wilson <[email protected]> Date: Fri Mar 11 21:25:48 2016 +0000
Quoted 7 lines
Quoted 3 lines
igt/drv_module_reload_basic: Rinse and repeat with addition module parameters
and apparently not intentionally.
It was, because the test wasn't behaving properly on my setup where I had disabled module unloading. If we can't unload the module, we have to SKIP the test.
I think a better check would be to lsmod | grep i915 and bail if that's false. Yeah that'll still fall over if you have unloading disabled, but either don't do that, or have some other check instead of rmmod i915 to assess whether unloading works.
And please don't hide such changes in another commit without even mentioning. i-g-t isn't free-for-all anymore, and if too much stuff slips through then we'd have to lock down commit process a lot. I kinda don't want that, since I still see plenty of benefits in making it simple to create more tests.
-Daniel
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx