linux-kernel - Re: [PATCH v3 0/2] dmaengine: idxd: IDXD pmu support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 23 Apr 2021 13:15:02 -0500
From:   "Zanussi, Tom" <tom.zanussi@...ux.intel.com>
To:     Vinod Koul <vkoul@...nel.org>
Cc:     peterz@...radead.org, acme@...nel.org, mingo@...nel.org,
        kan.liang@...ux.intel.com, dave.jiang@...el.com,
        tony.luck@...el.com, dan.j.williams@...el.com,
        linux-kernel@...r.kernel.org, dmaengine@...r.kernel.org
Subject: Re: [PATCH v3 0/2] dmaengine: idxd: IDXD pmu support

Hi Vinod,

On 4/23/2021 12:45 PM, Vinod Koul wrote:
> On 21-04-21, 16:04, Tom Zanussi wrote:
>> Hi,
>>
>> This is v3 of the IDXD pmu support patch, which addresses the comments
>> from Vinod:
>>
>>   - Removed the default line for INTEL_IDXD_PERFMON making it default 'n'
>>
>>   - Replaced #ifdef CONFIG_INTEL_IDXD_PERFMON with IS_ENABLED()
>>
>>   - Split the patch into two separate patches, the perfmon
>>     implementation and the code that uses it in the IDXD driver.
>>
>>   - Added a new file,
>>     Documentation/ABI/testing/sysfs-bus-event_source-devices-dsa that
>>     documents the new format and cpumask attributes, and added better
>>     comments for those in the code.
>>
>>   - Changed 'dogrp' to 'do_group' in perfmon_collect_events()
>>
>>   - Moved 'int idx' inside the loop in perfmon_validate_group() to the
>>     top of function.
>>
>>   - In perfmon_pmu_read_counter(), return ioread64() directly and get
>>     rid of cntrdata.
>>
>> I also fixed some erroneous code in perfmon_counter_overflow() that
>> because of my misreading of the spec caused unintended clearing of
>> wrong bits.  According to the spec you need to write 1 rather than 0
>> to an OVFSTATUS bit to clear it.
> 
> Applied, thanks
> 
> This conflicted with Daves patches, I managed to resolve, pls check the
> end result
> 

Thanks!

However, it looks like the new files in '[PATCH v3 1/2] dmaengine: idxd: Add
IDXD performance monitor support' didn't make it in, maybe didn't get 'added'
after resolving the conflicts...

+ create mode 100644 Documentation/ABI/testing/sysfs-bus-event_source-devices-dsa
+ create mode 100644 drivers/dma/idxd/perfmon.c
+ create mode 100644 drivers/dma/idxd/perfmon.h

Tom