Re: [PATCH v6] eal: allow worker lcore stacks to be allocated from hugepage memory

[Thomas Monjalon <thomas@monjalon.net>]

21/06/2022 16:52, Don Wallwork:
> On 6/21/2022 10:42 AM, Thomas Monjalon wrote:
> > 21/06/2022 14:31, Don Wallwork:
> >> On 6/21/2022 6:37 AM, Thomas Monjalon wrote:
> >>> 20/06/2022 10:35, David Marchand:
> >>>> On Tue, May 24, 2022 at 9:52 PM Don Wallwork <donw@xsightlabs.com> wrote:
> >>>>> Add support for using hugepages for worker lcore stack memory.  The
> >>>>> intent is to improve performance by reducing stack memory related TLB
> >>>>> misses and also by using memory local to the NUMA node of each lcore.
> >>>>> EAL option '--huge-worker-stack [stack-size-in-kbytes]' is added to allow
> >>>>> the feature to be enabled at runtime.  If the size is not specified,
> >>>>> the system pthread stack size will be used.
> >>>> - About the name of the option... I don't have a better name.
> >>>>
> >>>> Just want to highlight, that what this patch does is use the DPDK
> >>>> memory allocator for the stack memory.
> >>>> It happens that DPDK memory allocator is primarily used with
> >>>> hugepages, but this is not systematic for example with the "no-huge"
> >>>> mode of the DPDK memory allocator.
> >>>>
> >>>> IOW, in this patch current form, you can still run as:
> >>>>
> >>>> # dpdk-testpmd -c 3 --no-huge --huge-worker-stack=16 -m 40 -- etc...
> >>>>
> >>>> Opinions?
> >>> The name of the option should not include "huge".
> >>> What about "--worker-stack" ?
> >>> If disabled (equal zero), the workers should use the default stack memory.
> >>>
> >>>
> >> Wouldn't that have the potential to create confusion?  The point of this
> >> change is to allocate worker stacks from hugepages.  Removing huge
> >> from the option name could give the impression that the command is
> >> simply to control worker stack size.
> > It means if we control the worker stack size with a DPDK option,
> > DPDK memory will be used.
> > But we cannot force hugepage with this option.
> > Hugepage is not always available and it can be disabled in DPDK.
> 
> The command could be rejected if hugepages are not available.
> That's not in the patch currently, but can be added.

David, Anatoly, Dmitry, what do you think?