Date:   Wed, 20 Oct 2021 13:28:50 -0700
From:   Reinette Chatre <reinette.chatre@...el.com>
To:     Babu Moger <babu.moger@....com>, James Morse <james.morse@....com>,
        <x86@...nel.org>, <linux-kernel@...r.kernel.org>
CC:     Fenghua Yu <fenghua.yu@...el.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        H Peter Anvin <hpa@...or.com>,
        <shameerali.kolothum.thodi@...wei.com>,
        Jamie Iles <jamie@...iainc.com>,
        D Scott Phillips OS <scott@...amperecomputing.com>,
        <lcherian@...vell.com>, <bobo.shaobowang@...wei.com>,
        <tan.shaopeng@...itsu.com>
Subject: Re: [PATCH v2 17/23] x86/resctrl: Abstract __rmid_read()

Hi Babu,

On 10/20/2021 12:22 PM, Babu Moger wrote:
> On 10/20/21 1:15 PM, Reinette Chatre wrote:
>> On 10/19/2021 4:20 PM, Babu Moger wrote:
>>> On 10/1/21 11:02 AM, James Morse wrote:
>>>> __rmid_read() selects the specified eventid and returns the counter
>>>> value from the msr. The error handling is architecture specific, and
>>>> handled by the callers, rdtgroup_mondata_show() and __mon_event_count().
>>>>
>>>> Error handling should be handled by architecture specific code, as
>>>> a different architecture may have different requirements. MPAM's
>>>> counters can report that they are 'not ready', requiring a second
>>>> read after a short delay. This should be hidden from resctrl.
>>>>
>>>> Make __rmid_read() the architecture specific function for reading
>>>> a counter. Rename it resctrl_arch_rmid_read() and move the error
>>>> handling into it.
>>>>
>>>> Signed-off-by: James Morse <james.morse@....com>
>>>> ---
>>>> Changes since v1:
>>>>    * Return EINVAL from the impossible case in __mon_event_count() instead
>>>>      of an x86 hardware specific value.
>>>> ---
>>>>    arch/x86/kernel/cpu/resctrl/ctrlmondata.c |  4 +--
>>>>    arch/x86/kernel/cpu/resctrl/internal.h    |  2 +-
>>>>    arch/x86/kernel/cpu/resctrl/monitor.c     | 42 +++++++++++++++--------
>>>>    include/linux/resctrl.h                   |  1 +
>>>>    4 files changed, 31 insertions(+), 18 deletions(-)
>>>>
>>>> diff --git a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>>>> b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>>>> index 25baacd331e0..c8ca7184c6d9 100644
>>>> --- a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>>>> +++ b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>>>> @@ -579,9 +579,9 @@ int rdtgroup_mondata_show(struct seq_file *m, void
>>>> *arg)
>>>>          mon_event_read(&rr, r, d, rdtgrp, evtid, false);
>>>>    -    if (rr.val & RMID_VAL_ERROR)
>>>> +    if (rr.err == -EIO)
>>>>            seq_puts(m, "Error\n");
>>>> -    else if (rr.val & RMID_VAL_UNAVAIL)
>>>> +    else if (rr.err == -EINVAL)
>>>>            seq_puts(m, "Unavailable\n");
>>>>        else
>>>>            seq_printf(m, "%llu\n", rr.val * hw_res->mon_scale);
>>>
>>> This patch breaks the earlier fix
>>> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.kernel.org%2Fpub%2Fscm%2Flinux%2Fkernel%2Fgit%2Ftorvalds%2Flinux.git%2Fcommit%2F%3Fh%3Dv5.15-rc6%26id%3D064855a69003c24bd6b473b367d364e418c57625&amp;data=04%7C01%7Cbabu.moger%40amd.com%7C85219a5827114935cdaa08d993f59fa0%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637703505420472920%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=yP8awDgGGZ%2BWj5ZItdTNJItTVuK828yGnibwq%2BrVaf0%3D&amp;reserved=0
>>>
>>>
>>> When the user reads the events on the default monitoring group with
>>> multiple subgroups, the events on all subgroups are consolidated
>>> together. In case if the last rmid read was resulted in error then whole
>>> group will be reported as error. The err field needs to be cleared.
>>>
>>> Please add this patch to clear the error.
>>>
>>> diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c
>>> b/arch/x86/kernel/cpu/resctrl/monitor.c
>>> index 14bc843043da..0e4addf237ec 100644
>>> --- a/arch/x86/kernel/cpu/resctrl/monitor.c
>>> +++ b/arch/x86/kernel/cpu/resctrl/monitor.c
>>> @@ -444,6 +444,8 @@ void mon_event_count(void *info)
>>>           /* Report error if none of rmid_reads are successful */
>>>           if (ret_val)
>>>                   rr->val = ret_val;
>>> +       else
>>> +               rr->err  = 0;
>>>    }
>>>
>>>    /*
>>>
>>
>> Good catch, thank you.
>>
>> Even so, I do not think mon_event_count()'s usage of __mon_event_count()
>> was taken into account by this patch and needs a bigger rework than the
>> above fixup. For example, if I understand correctly ret_val is the error
>> and rr->val no longer expected to contain the error after this patch. So
>> keeping that assignment to rr->val is not correct.
> 
> Yes. You are right. rr->val is not expected to contain the error.
> Hopefully, this should help.
> 
> diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c
> b/arch/x86/kernel/cpu/resctrl/monitor.c
> index 14bc843043da..105d972cc511 100644
> --- a/arch/x86/kernel/cpu/resctrl/monitor.c
> +++ b/arch/x86/kernel/cpu/resctrl/monitor.c
> @@ -441,9 +441,9 @@ void mon_event_count(void *info)
>                  }
>          }
> 
> -       /* Report error if none of rmid_reads are successful */
> -       if (ret_val)
> -               rr->val = ret_val;
> +       /* Clear the error if at least one of the rmid reads succeed */
> +       if (ret_val == 0)
> +               rr->err = 0;
>   }
> 
>   /*
> 

Yes, this looks good. If the first __mon_event_count() succeeds but a 
following one fails then the data still needs to be reported so the 
error code needs to be fixed up afterwards and cannot be done inside 
__mon_event_count(). Thank you very much.

Reinette

Powered by blists - more mailing lists