diff mbox series

[blktests] srp: skip test if scsi_transport_srp module is loaded and in use

Message ID 20250201184021.278437-1-nilay@linux.ibm.com (mailing list archive)
State New
Headers show
Series [blktests] srp: skip test if scsi_transport_srp module is loaded and in use | expand

Commit Message

Nilay Shroff Feb. 1, 2025, 6:40 p.m. UTC
The srp/* tests requires exclusive access to scsi_transport_srp
module. Running srp/* tests would definitely fail if the test can't
get exclusive access of scsi_transport_srp module as shown below:

$ lsmod | grep scsi_transport_srp
scsi_transport_srp    327680  1 ibmvscsi

$ ./check srp/001
srp/001 (Create and remove LUNs)                             [failed]
    runtime    ...  0.249s
tests/srp/rc: line 263: /sys/class/srp_remote_ports/port-0:1/delete: Permission denied
tests/srp/rc: line 263: /sys/class/srp_remote_ports/port-0:1/delete: Permission denied
modprobe: FATAL: Module scsi_transport_srp is in use.
error: Invalid argument
error: Invalid argument

So if the scsi_transport_srp module is loaded and in use then skip
running srp/* tests.

Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
---
 common/rc    | 11 +++++++++++
 tests/srp/rc |  1 +
 2 files changed, 12 insertions(+)

Comments

Shinichiro Kawasaki Feb. 4, 2025, 11:55 a.m. UTC | #1
CC+: Bart,

On Feb 02, 2025 / 00:10, Nilay Shroff wrote:
> The srp/* tests requires exclusive access to scsi_transport_srp
> module. Running srp/* tests would definitely fail if the test can't
> get exclusive access of scsi_transport_srp module as shown below:
> 
> $ lsmod | grep scsi_transport_srp
> scsi_transport_srp    327680  1 ibmvscsi
> 
> $ ./check srp/001
> srp/001 (Create and remove LUNs)                             [failed]
>     runtime    ...  0.249s
> tests/srp/rc: line 263: /sys/class/srp_remote_ports/port-0:1/delete: Permission denied
> tests/srp/rc: line 263: /sys/class/srp_remote_ports/port-0:1/delete: Permission denied
> modprobe: FATAL: Module scsi_transport_srp is in use.
> error: Invalid argument
> error: Invalid argument
> 
> So if the scsi_transport_srp module is loaded and in use then skip
> running srp/* tests.

Thanks. I think this is a good improvement.

> 
> Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
> ---
>  common/rc    | 11 +++++++++++
>  tests/srp/rc |  1 +
>  2 files changed, 12 insertions(+)
> 
> diff --git a/common/rc b/common/rc
> index bcb215d..73e0b9a 100644
> --- a/common/rc
> +++ b/common/rc
> @@ -78,6 +78,17 @@ _have_module() {
>  	return 0
>  }
>  
> +_have_module_not_in_use() {
> +       _have_module "$1" || return
> +
> +       if [ -d "/sys/module/$1" ]; then
> +               refcnt="$(cat /sys/module/$1/refcnt)"
> +               if [ "$refcnt" -ne "0" ]; then
> +                       SKIP_REASONS+=("module $1 is in use")
> +               fi
> +       fi
> +}

Spaces are used for indents. Please use tabs instead.

Nit: "refcnt" is not declared as a local variable. Let's declare it. Or, it
would be the better to avoid "refcnt". Instead, $(< /sys/module/$1/refcnt)
can be used as follows. Either way is fine for me.

       if [ -d "/sys/module/$1" ] && (($(</sys/module/"$1"/refcnt) != 0)); then
               SKIP_REASONS+=("module $1 is in use")
       fi
Bart Van Assche Feb. 4, 2025, 6 p.m. UTC | #2
On 2/1/25 10:40 AM, Nilay Shroff wrote:
> +_have_module_not_in_use() {

The name "_have_module_not_in_use()" sounds weird to me. Wouldn't
_module_not_in_use() be a better name?

Thanks,

Bart.
Nilay Shroff Feb. 5, 2025, 1:27 p.m. UTC | #3
On 2/4/25 5:25 PM, Shinichiro Kawasaki wrote:
> CC+: Bart,
> 
> On Feb 02, 2025 / 00:10, Nilay Shroff wrote:
>> The srp/* tests requires exclusive access to scsi_transport_srp
>> module. Running srp/* tests would definitely fail if the test can't
>> get exclusive access of scsi_transport_srp module as shown below:
>>
>> $ lsmod | grep scsi_transport_srp
>> scsi_transport_srp    327680  1 ibmvscsi
>>
>> $ ./check srp/001
>> srp/001 (Create and remove LUNs)                             [failed]
>>     runtime    ...  0.249s
>> tests/srp/rc: line 263: /sys/class/srp_remote_ports/port-0:1/delete: Permission denied
>> tests/srp/rc: line 263: /sys/class/srp_remote_ports/port-0:1/delete: Permission denied
>> modprobe: FATAL: Module scsi_transport_srp is in use.
>> error: Invalid argument
>> error: Invalid argument
>>
>> So if the scsi_transport_srp module is loaded and in use then skip
>> running srp/* tests.
> 
> Thanks. I think this is a good improvement.
> 
>>
>> Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
>> ---
>>  common/rc    | 11 +++++++++++
>>  tests/srp/rc |  1 +
>>  2 files changed, 12 insertions(+)
>>
>> diff --git a/common/rc b/common/rc
>> index bcb215d..73e0b9a 100644
>> --- a/common/rc
>> +++ b/common/rc
>> @@ -78,6 +78,17 @@ _have_module() {
>>  	return 0
>>  }
>>  
>> +_have_module_not_in_use() {
>> +       _have_module "$1" || return
>> +
>> +       if [ -d "/sys/module/$1" ]; then
>> +               refcnt="$(cat /sys/module/$1/refcnt)"
>> +               if [ "$refcnt" -ne "0" ]; then
>> +                       SKIP_REASONS+=("module $1 is in use")
>> +               fi
>> +       fi
>> +}
> 
> Spaces are used for indents. Please use tabs instead.
> 
Grrr.. I think my editor inserted spaces instead of tabs, I would fix this
in the next version of the patch.

> Nit: "refcnt" is not declared as a local variable. Let's declare it. Or, it
> would be the better to avoid "refcnt". Instead, $(< /sys/module/$1/refcnt)
> can be used as follows. Either way is fine for me.
> 
>        if [ -d "/sys/module/$1" ] && (($(</sys/module/"$1"/refcnt) != 0)); then
>                SKIP_REASONS+=("module $1 is in use")
>        fi
Okay, I'd make "refcnt" local and update.

Thanks,
--Nilay
Nilay Shroff Feb. 5, 2025, 1:29 p.m. UTC | #4
On 2/4/25 11:30 PM, Bart Van Assche wrote:
> On 2/1/25 10:40 AM, Nilay Shroff wrote:
>> +_have_module_not_in_use() {
> 
> The name "_have_module_not_in_use()" sounds weird to me. Wouldn't
> _module_not_in_use() be a better name?
> 
Yeah _module_not_in_use() sounds better. I will incorporate this
in the next patch.

Thanks,
--Nilay
diff mbox series

Patch

diff --git a/common/rc b/common/rc
index bcb215d..73e0b9a 100644
--- a/common/rc
+++ b/common/rc
@@ -78,6 +78,17 @@  _have_module() {
 	return 0
 }
 
+_have_module_not_in_use() {
+       _have_module "$1" || return
+
+       if [ -d "/sys/module/$1" ]; then
+               refcnt="$(cat /sys/module/$1/refcnt)"
+               if [ "$refcnt" -ne "0" ]; then
+                       SKIP_REASONS+=("module $1 is in use")
+               fi
+       fi
+}
+
 _have_module_param() {
 	 _have_driver "$1" || return
 
diff --git a/tests/srp/rc b/tests/srp/rc
index 85bd1dd..1bc7b20 100755
--- a/tests/srp/rc
+++ b/tests/srp/rc
@@ -61,6 +61,7 @@  group_requires() {
 	_have_module scsi_debug
 	_have_module target_core_iblock
 	_have_module target_core_mod
+	_have_module_not_in_use scsi_transport_srp
 
 	for p in mkfs.ext4 mkfs.xfs multipath multipathd pidof rdma \
 		 sg_reset fio; do