Re: [PATCH bpf-next v1 RESEND 1/5] vmalloc: introduce vmalloc_exec, vfree_exec, and vcopy_exec

[Luis Chamberlain <mcgrof@kernel.org>]

On Thu, Nov 03, 2022 at 09:19:25PM +0000, Edgecombe, Rick P wrote:
> On Thu, 2022-11-03 at 11:59 -0700, Luis Chamberlain wrote:
> > > > Mike Rapoport had presented about the Direct map fragmentation
> > > > problem
> > > > at Plumbers 2021 [0], and clearly mentioned modules / BPF /
> > > > ftrace /
> > > > kprobes as possible sources for this. Then Xing Zhengjun's 2021
> > > > performance
> > > > evaluation on whether using 2M/1G pages aggressively for the
> > > > kernel direct map
> > > > help performance [1] ends up generally recommending huge pages.
> > > > The work by Xing
> > > > though was about using huge pages *alone*, not using a strategy
> > > > such as in the
> > > > "bpf prog pack" to share one 2 MiB huge page for *all* small eBPF
> > > > programs,
> > > > and that I think is the real golden nugget here.
> > > > 
> > > > I contend therefore that the theoretical reduction of iTLB misses
> > > > by using
> > > > huge pages for "bpf prog pack" is not what gets your systems to
> > > > perform
> > > > somehow better. It should be simply that it reduces fragmentation
> > > > and
> > > > *this* generally can help with performance long term. If this is
> > > > accurate
> > > > then let's please separate the two aspects to this.
> > > 
> > > The direct map fragmentation is the reason for higher TLB miss
> > > rate, both
> > > for iTLB and dTLB.
> > 
> > OK so then whatever benchmark is running in tandem as eBPF JIT is
> > hammered
> > should *also* be measured with perf for iTLB and dTLB. ie, the patch
> > can
> > provide such results as a justifications.
> 
> Song had done some tests on the old prog pack version that to me seemed
> to indicate most (or possibly all) of the benefit was direct map
> fragmentation reduction.

Matches my observations but I also provided quite a bit of hints as to
*why* I think that is. I suggested lib/test_kmod.c as an example beefy
multithreaded selftests which really kicks the hell out of the kernel
with whatever crap you want to run. That is precicely how I uncovered
some odd kmod bug lingering for years.

> This was surprised me, since 2MB kernel text
> has shown to be beneficial.
> 
> Otherwise +1 to all these comments. This should be clear about what the
> benefits are. I would add, that this is also much nicer about TLB
> shootdowns than the existing way of loading text and saves some memory.
> 
> So I think there are sort of four areas of improvements:
> 1. Direct map fragmentation reduction (dTLB miss improvements).

The dTLB gains should be on the benchmark which runs in tandem to the
ebpf-JIT-monster-selftest, not on the ebpf-JIT-monster-selftest, right?

> This
> sort of does it as a side effect in this series, and the solution Mike
> is talking about is a more general, probably better one.
> 2. 2MB mapped JITs. This is the iTLB side. I think this is a decent
> solution for this, but surprisingly it doesn't seem to be useful for
> JITs. (modules testing TBD)

Yes I'm super eager to get this tested. In fact I wonder if one can
boot Linux with less memory too...

> 3. Loading text to reused allocation with per-cpu mappings. This
> reduces TLB shootdowns, which are a short term load and teardown time
> performance drag. My understanding is this is more of a problem on
> bigger systems with many CPUs. This series does a decent job at this,
> but the solution is not compatible with modules. Maybe ok since modules
> don't load as often as JITs.

There are some tests like fstests which make heavy use of module
removal. But as a side effect, indeed I like to reboot to have
a fresh system before running fstests. I guess fstests should run
with a heavily fragmented memory too as a side corner case thing.

> 4. Having BPF progs share pages. This saves memory. This series could
> probably easily get a number for how much.

Once that does hit modules / kprobes / ftrace, the impact is much
much greater obviously.

  Luis