[PATCH RESEND v11 0/8] proc: modernize proc to support multiple private instances

[Eric W. Biederman]

Alexey Gladkov <gladkov.alexey at gmail.com> writes:

> On Thu, Apr 09, 2020 at 07:59:47AM -0500, Eric W. Biederman wrote:
>> Alexey Gladkov <gladkov.alexey at gmail.com> writes:
>> 
>> > Preface:
>> > --------
>> > This is patchset v11 to modernize procfs and make it able to support multiple
>> > private instances per the same pid namespace.
>> >
>> > This patchset can be applied on top of:
>> >
>> > git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace.git
>> > 4b871ce26ab2
>> 
>> 
>> 
>> Why the resend?
>> 
>> Nothing happens until the merge window closes with the release of -rc1
>> (almost certainly on this coming Sunday).  I goofed and did not act on
>> this faster, and so it is my fault this did not make it into linux-next
>> before the merge window.  But I am not going to rush this forward.
>> 
>> 
>> 
>> You also ignored my review and have not even descibed why it is safe
>> to change the type of a filesystem parameter.
>> 
>> -	fsparam_u32("hidepid",	Opt_hidepid),
>> +	fsparam_string("hidepid",	Opt_hidepid),
>> 
>> 
>> Especially in light of people using fsconfig(fd, FSCONFIG_SET_...);
>> 
>> All I need is someone to point out that fsparam_u32 does not use
>> FSCONFIG_SET_BINARY, but FSCONFIG_SET_STRING.
>
> I decided to resend again because I was not sure that the previous
> patchset was not lost. I also wanted to ask David to review and explain
> about the new API. I in any case did not ignore your question about
> changing the type of the parameter.
>
> I guess I was wrong when I sent the whole patchset again. Sorry.
>
>> My apologies for being grumpy but this feels like you are asking me to
>> go faster when it is totally inappropriate to do so, while busily
>> ignoring my feedback.
>> 
>> I think this should happen.  But I can't do anything until after -rc1.
>
> I think you misunderstood me. I didn't mean to rush you.

It looks like.  My apologies for the misunderstanding then.

Eric