2019-05-08 17:21:23 +02:00
|
|
|
.. SPDX-License-Identifier: GPL-2.0
|
|
|
|
|
|
|
|
=======
|
|
|
|
The TLB
|
|
|
|
=======
|
|
|
|
|
2014-07-31 17:41:01 +02:00
|
|
|
When the kernel unmaps or modified the attributes of a range of
|
|
|
|
memory, it has two choices:
|
2019-05-08 17:21:23 +02:00
|
|
|
|
2014-07-31 17:41:01 +02:00
|
|
|
1. Flush the entire TLB with a two-instruction sequence. This is
|
|
|
|
a quick operation, but it causes collateral damage: TLB entries
|
|
|
|
from areas other than the one we are trying to flush will be
|
|
|
|
destroyed and must be refilled later, at some cost.
|
|
|
|
2. Use the invlpg instruction to invalidate a single page at a
|
2016-07-01 05:46:01 +02:00
|
|
|
time. This could potentially cost many more instructions, but
|
2014-07-31 17:41:01 +02:00
|
|
|
it is a much more precise operation, causing no collateral
|
|
|
|
damage to other TLB entries.
|
|
|
|
|
|
|
|
Which method to do depends on a few things:
|
2019-05-08 17:21:23 +02:00
|
|
|
|
2014-07-31 17:41:01 +02:00
|
|
|
1. The size of the flush being performed. A flush of the entire
|
|
|
|
address space is obviously better performed by flushing the
|
|
|
|
entire TLB than doing 2^48/PAGE_SIZE individual flushes.
|
|
|
|
2. The contents of the TLB. If the TLB is empty, then there will
|
|
|
|
be no collateral damage caused by doing the global flush, and
|
|
|
|
all of the individual flush will have ended up being wasted
|
|
|
|
work.
|
|
|
|
3. The size of the TLB. The larger the TLB, the more collateral
|
|
|
|
damage we do with a full flush. So, the larger the TLB, the
|
2016-07-01 05:46:01 +02:00
|
|
|
more attractive an individual flush looks. Data and
|
2014-07-31 17:41:01 +02:00
|
|
|
instructions have separate TLBs, as do different page sizes.
|
|
|
|
4. The microarchitecture. The TLB has become a multi-level
|
|
|
|
cache on modern CPUs, and the global flushes have become more
|
|
|
|
expensive relative to single-page flushes.
|
|
|
|
|
|
|
|
There is obviously no way the kernel can know all these things,
|
|
|
|
especially the contents of the TLB during a given flush. The
|
|
|
|
sizes of the flush will vary greatly depending on the workload as
|
|
|
|
well. There is essentially no "right" point to choose.
|
|
|
|
|
|
|
|
You may be doing too many individual invalidations if you see the
|
|
|
|
invlpg instruction (or instructions _near_ it) show up high in
|
|
|
|
profiles. If you believe that individual invalidations being
|
2019-05-08 17:21:23 +02:00
|
|
|
called too often, you can lower the tunable::
|
2014-07-31 17:41:01 +02:00
|
|
|
|
2014-08-08 09:49:55 +02:00
|
|
|
/sys/kernel/debug/x86/tlb_single_page_flush_ceiling
|
2014-07-31 17:41:01 +02:00
|
|
|
|
|
|
|
This will cause us to do the global flush for more cases.
|
|
|
|
Lowering it to 0 will disable the use of the individual flushes.
|
|
|
|
Setting it to 1 is a very conservative setting and it should
|
|
|
|
never need to be 0 under normal circumstances.
|
|
|
|
|
|
|
|
Despite the fact that a single individual flush on x86 is
|
2019-05-08 17:21:23 +02:00
|
|
|
guaranteed to flush a full 2MB [1]_, hugetlbfs always uses the full
|
2014-07-31 17:41:01 +02:00
|
|
|
flushes. THP is treated exactly the same as normal memory.
|
|
|
|
|
|
|
|
You might see invlpg inside of flush_tlb_mm_range() show up in
|
|
|
|
profiles, or you can use the trace_tlb_flush() tracepoints. to
|
|
|
|
determine how long the flush operations are taking.
|
|
|
|
|
|
|
|
Essentially, you are balancing the cycles you spend doing invlpg
|
|
|
|
with the cycles that you spend refilling the TLB later.
|
|
|
|
|
|
|
|
You can measure how expensive TLB refills are by using
|
2019-05-08 17:21:23 +02:00
|
|
|
performance counters and 'perf stat', like this::
|
2014-07-31 17:41:01 +02:00
|
|
|
|
2019-05-08 17:21:23 +02:00
|
|
|
perf stat -e
|
|
|
|
cpu/event=0x8,umask=0x84,name=dtlb_load_misses_walk_duration/,
|
|
|
|
cpu/event=0x8,umask=0x82,name=dtlb_load_misses_walk_completed/,
|
|
|
|
cpu/event=0x49,umask=0x4,name=dtlb_store_misses_walk_duration/,
|
|
|
|
cpu/event=0x49,umask=0x2,name=dtlb_store_misses_walk_completed/,
|
|
|
|
cpu/event=0x85,umask=0x4,name=itlb_misses_walk_duration/,
|
|
|
|
cpu/event=0x85,umask=0x2,name=itlb_misses_walk_completed/
|
2014-07-31 17:41:01 +02:00
|
|
|
|
|
|
|
That works on an IvyBridge-era CPU (i5-3320M). Different CPUs
|
|
|
|
may have differently-named counters, but they should at least
|
|
|
|
be there in some form. You can use pmu-tools 'ocperf list'
|
|
|
|
(https://github.com/andikleen/pmu-tools) to find the right
|
|
|
|
counters for a given CPU.
|
|
|
|
|
2019-05-08 17:21:23 +02:00
|
|
|
.. [1] A footnote in Intel's SDM "4.10.4.2 Recommended Invalidation"
|
2014-07-31 17:41:01 +02:00
|
|
|
says: "One execution of INVLPG is sufficient even for a page
|
|
|
|
with size greater than 4 KBytes."
|