On 1/9/23 15:25, Luck, Tony wrote:
>>> All the QoS(or RDT) features are visible so far. If we make them visible,
>>> users can easily figure out if this specific feature is supported or not.
>> What would be the actual, real-life use case where the presence of those flags
>> in /proc/cpuinfo is really needed?
> It feels like the old "rule" was "make it visible in /proc/cpuid" unless there was some
> good reason NOT to do it. But that has resulted in the "flags" line getting ridiculously
> long and hard for humans to read (141 fields with 926 bytes on my Skylake, more on
> more modern CPUs).
>
> For RDT I don't see a lot of value in knowing that a feature is present ... all of them
> have parameters on how many things they can control/monitor ... so you have to
> either go parse the CPUID leaves, or just mount /sys/fs/resctrl and look in the "info"
> directory to get the extra information you need to do anything with RDT.
>
> I don't know if we'd break anything if we dropped:
>
> cat_l3 cdp_l3 mba cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local
>
> from /proc/cpuinfo.
>
> Perhaps the "rule" should be written in Documentation/{somewhere}?
Actually, these feature bits are referred in Documentation/x86/resctrl.rst
This feature is enabled by the CONFIG_X86_CPU_RESCTRL and the x86
/proc/cpuinfo
flag bits:
===============================================
================================
RDT (Resource Director Technology) Allocation "rdt_a"
CAT (Cache Allocation Technology) "cat_l3", "cat_l2"
CDP (Code and Data Prioritization) "cdp_l3", "cdp_l2"
CQM (Cache QoS Monitoring) "cqm_llc", "cqm_occup_llc"
MBM (Memory Bandwidth Monitoring) "cqm_mbm_total",
"cqm_mbm_local"
MBA (Memory Bandwidth Allocation) "mba"
SMBA (Slow Memory Bandwidth Allocation) "smba"
BMEC (Bandwidth Monitoring Event Configuration) "bmec"
===============================================
================================
So, if we remove them, we need to update here also.
Thanks
Babu
