Re: [PATCH net] net/mlx5e: Do not recycle pages from emergency reserve

On Sat, Jan 21, 2017 at 11:12 AM, kernel netdev [off-list ref] wrote:

Den 21. jan. 2017 7.10 PM skrev "Tom Herbert" [off-list ref]:

On Thu, Jan 19, 2017 at 11:14 AM, Saeed Mahameed
[off-list ref] wrote:

quoted

On Thu, Jan 19, 2017 at 9:03 AM, Eric Dumazet [off-list ref]
wrote:

quoted

From: Eric Dumazet <edumazet@google.com>

A driver using dev_alloc_page() must not reuse a page allocated from
emergency memory reserve.

Otherwise all packets using this page will be immediately dropped,
unless for very specific sockets having SOCK_MEMALLOC bit set.

This issue might be hard to debug, because only a fraction of received
packets would be dropped.

Hi Eric,

When you say reuse, you mean point to the same page from several SKBs ?

Because in our page cache implementation we don't reuse pages that
already passed to the stack,
we just keep them in the page cache until the ref count drop back to
one, so we recycle them (i,e they will be re-used only when no one
else is using them).

Saeed,

Speaking of the mlx page cache can we remove this or a least make it
optional to use. It is another example of complex functionality being
put into drivers that makes things like backports more complicated and
provide at best some marginal value. In the case of the mlx5e cache
code the results from pktgen really weren't very impressive in the
first place. Also, the cache suffers from HOL blocking where we can
block the whole cache due to an outstanding reference on just one page
(something that you wouldn't see in pktgen but is likely to happen in
real applications).


(Send from phone in car)

To Tom, have you measured the effect of this page cache? Before claiming it
is ineffective.

No, I have not. TBH, I have most of the past few weeks trying to debug
a backport of the code from 4.9 to 4.6. Until we have a working
backport performance is immaterial for our purposes. Unfortunately, we
are seeing some issues: the checksum faults I described previously and
crashes on bad page refcns which are presumably being caused by the
logic in RX buffer processing. This is why I am now having to dissect
the code and trying to disable things like the page cache that are not
essential functionality.

In any case the HOL blocking issue is obvious from reading the code
which and implies bimodal behavior-- we don't need a test for that to
know it's a bad as we've see the bad effects of that in many other
contexts.

My previous measurements show approx 20℅ speedup on a UDP test with delivery
to remote CPU.

Removing the cache would of cause be a good usecase for speeding up the page
allocator (PCP). Which Mel Gorman and me are working on. AFAIK current page
order0 cost 240 cycles. Mel have reduced til to 180, and without NUMA 150
cycles. And with bulking this can be amortized to 80 cycles.

That would be great. If only I had a nickel for every time someone
started working on a driver and came the conclusion that they need to
do a custom memory allocator because the kernel allocator is so
inefficient!

Tom

--Jesper