Re: [PATCH 2/4] nvme-multipath: add the NVME_MULTIPATH_PARAM config option

[John Meneghini <jmeneghi@redhat.com>]

On 2/28/25 1:28 AM, Nilay Shroff wrote: 
> On 2/28/25 8:55 AM, John Meneghini wrote:
>> diff --git a/drivers/nvme/host/Kconfig b/drivers/nvme/host/Kconfig
>> index 91b0346ce65a..c4251504f201 100644
>> --- a/drivers/nvme/host/Kconfig
>> +++ b/drivers/nvme/host/Kconfig
>> @@ -28,6 +28,21 @@ config NVME_MULTIPATH
>>   
>>   	  If unsure, say Y.
>>   
>> +config NVME_MULTIPATH_PARAM
>> +	bool "NVMe multipath param"
>> +	depends on NVME_CORE && NVME_MULTIPATH
>> +	help
>> +	  This option enables configurable support for multipath access with
>> +	  NVMe subsystems. If this option is enabled NVMe multipath support is
>> +	  configured by the nvme core module parameter named "multipath". If
>> +	  this option is disabled the nvme core module "multipath" parameter
>> +	  is removed and support for NVMe multipath access can not be
>> +	  configured. When this option is disabled a single /dev/nvmeXnY
>> +	  device entry will be seen for each NVMe namespace, even if the
>> +	  namespace is accessible through multiple controllers.
>> +
>> +	  If unsure, say Y.
>> +
> If we want to make NVME_MULTIPATH_PARAM default on then I think we need to add
> "default y" under config NVME_MULTIPATH_PARAM.

OK. I've tested all of the config options.

make mod2noconfig
make allyesconfig
make allmodconfig
make oldconfig

And is all seems to be working correctly, but I'll add the "default y"
as you've suggested.

>>   config NVME_VERBOSE_ERRORS
>>   	bool "NVMe verbose error reporting"
>>   	depends on NVME_CORE
>> diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
>> index 2a7635565083..4536ad5fbb82 100644
>> --- a/drivers/nvme/host/multipath.c
>> +++ b/drivers/nvme/host/multipath.c
>> @@ -10,10 +10,11 @@
>>   #include "nvme.h"
>>   
>>   bool multipath = true;
>> +#ifdef NVME_MULTIPATH_PARAM
> 
> Shouldn't it be CONFIG_NVME_MULTIPATH_PARAM instead of NVME_MULTIPATH_PARAM?

Oops.  As you can tell, I haven't tested this yet.  I'll fix this up and test these
changes before sending a version 2 patch.

/John
  
> Thanks,
> --Nilay
>