From: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
To: Paul Mackerras <paulus@ozlabs.org>
Cc: benh@kernel.crashing.org, mpe@ellerman.id.au,
linuxppc-dev@lists.ozlabs.org
Subject: Re: [PATCH 00/16] Remove hash page table slot tracking from linux PTE
Date: Fri, 27 Oct 2017 10:57:13 +0530 [thread overview]
Message-ID: <adf2d270-29df-701f-206e-0d8a35084e47@linux.vnet.ibm.com> (raw)
In-Reply-To: <20171027043430.GA27483@fergus.ozlabs.ibm.com>
On 10/27/2017 10:04 AM, Paul Mackerras wrote:
> On Fri, Oct 27, 2017 at 09:38:17AM +0530, Aneesh Kumar K.V wrote:
>> Hi,
>>
>> With hash translation mode we always tracked the hash pte slot details in linux page table.
>> This occupied space in the linux page table and also limitted our ability to support
>> linux features that require additional PTE bits. This series attempt to lift this
>> limitation by not tracking slot number in linux page table. We still track slot details
>> w.r.t Transparent Hugepage entries because an invalidate there requires us to go through
>> all the 256 hash pte slots. So tracking whether hash page table entry is valid helps us in
>> avoiding a lot of hcalls there. With THP entries we don't keep slot details in the primary
>> linux page table entry but in the second half of page table. Hence tracking slot details
>> for THP doesn't take up space in PTE.
>>
>> Even though we don't track slot, for removing/updating hash page table entry, PAPR hcalls expect
>> hash page table slot details. On pseries we find slot using H_READ hcall using H_READ_4 flags.
>> This implies an additional 2 hcalls in the updatepp and remove paths. The patch series also
>> attempt to limit the impact of this by adding new hcalls that does remove/update of hash page table
>> entry using hash value instead of hash page table slot.
>>
>> Below is the performance numbers observed when running a workload that does the below sequence
>>
>> for(5000) {
>> mmap(128M)
>> touch every page of 2048 page
>> munmap()
>> }
>>
>> The test is run with address randomization off, swap disabled in both host and guest.
>>
>>
>> |------------+----------+---------------+--------------------------+-----------------------|
>> | iterations | platform | without patch | With series and no hcall | With series and hcall |
>> |------------+----------+---------------+--------------------------+-----------------------|
>> | 1 | powernv | | 50.818343 | |
>> | 2 | powernv | | 50.744123 | |
>> | 3 | powernv | | 50.721603 | |
>> | 4 | powernv | | 50.739922 | |
>> | 5 | powernv | | 50.638555 | |
>> | 1 | powernv | 51.388249 | | |
>> | 2 | powernv | 51.789701 | | |
>> | 3 | powernv | 52.240394 | | |
>> | 4 | powernv | 51.432255 | | |
>> | 5 | powernv | 51.392947 | | |
>> |------------+----------+---------------+--------------------------+-----------------------|
>> | 1 | pseries | | | 123.154394 |
>> | 2 | pseries | | | 122.253956 |
>> | 3 | pseries | | | 117.666344 |
>> | 4 | pseries | | | 117.681479 |
>> | 5 | pseries | | | 117.735808 |
>> | 1 | pseries | | 119.424940 | |
>> | 2 | pseries | | 117.663078 | |
>> | 3 | pseries | | 118.345584 | |
>> | 4 | pseries | | 119.620934 | |
>> | 5 | pseries | | 119.463185 | |
>> | 1 | pseries | 122.810867 | | |
>> | 2 | pseries | 115.760801 | | |
>> | 3 | pseries | 115.257030 | | |
>> | 4 | pseries | 116.617884 | | |
>> | 5 | pseries | 117.247036 | | |
>> |------------+----------+---------------+--------------------------+-----------------------|
>>
>
> How do we interpret these numbers? Are they times, or speed? Is
> larger better or worse?
Sorry for not including the details. They are time in seconds. Test case
is a modified mmap_bench included in powerpc/selftest.
>
> Can you give us the mean and standard deviation for each set of 5
> please?
>
powernv without patch
median= 51.432255
stdev = 0.370835
with patch
median = 50.739922
stdev = 0.06419662
pseries without patch
median = 116.617884
stdev = 3.04531023
with patch no hcall
median = 119.42494
stdev = 0.85874552
with patch and hcall
median = 117.735808
stdev = 2.7624151
-aneesh
next prev parent reply other threads:[~2017-10-27 5:27 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-10-27 4:08 [PATCH 00/16] Remove hash page table slot tracking from linux PTE Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 01/16] powerpc/mm/hash: Remove the superfluous bitwise operation when find hpte group Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 02/16] powerpc/mm: Update native_hpte_find to return hash pte Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 03/16] powerpc/pseries: Update hpte find helper to take hash value Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 04/16] powerpc/mm: Add hash invalidate callback Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 05/16] powerpc/mm: use hash_invalidate for __kernel_map_pages() Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 06/16] powerpc/mm: Switch flush_hash_range to not use slot Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 07/16] powerpc/mm: Add hash updatepp callback Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 08/16] powerpc/mm/hash: Don't track hash pte slot number in linux page table Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 09/16] powerpc/mm: Add new firmware feature HASH API Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 10/16] powerpc/kvm/hash: Implement HASH_REMOVE hcall Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 11/16] powerpc/kvm/hash: Implement HASH_PROTECT hcall Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 12/16] powerpc/kvm/hash: Implement HASH_BULK_REMOVE hcall Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 13/16] powerpc/mm/pseries: Use HASH_PROTECT hcall in guest Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 14/16] powerpc/mm/pseries: Use HASH_REMOVE " Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 15/16] powerpc/mm/pseries: Move slot based bulk remove to helper Aneesh Kumar K.V
2017-10-27 4:08 ` [PATCH 16/16] powerpc/mm/pseries: Use HASH_BULK_REMOVE hcall in guest Aneesh Kumar K.V
2017-10-27 4:34 ` [PATCH 00/16] Remove hash page table slot tracking from linux PTE Paul Mackerras
2017-10-27 5:27 ` Aneesh Kumar K.V [this message]
2017-10-27 5:41 ` Paul Mackerras
2017-10-30 7:57 ` Aneesh Kumar K.V
2017-10-30 11:49 ` Aneesh Kumar K.V
2017-10-30 13:14 ` Aneesh Kumar K.V
2017-10-30 13:49 ` Aneesh Kumar K.V
2017-11-21 8:41 ` Aneesh Kumar K.V
2017-10-28 22:35 ` Ram Pai
2017-10-29 14:05 ` Aneesh Kumar K.V
2017-10-29 22:04 ` Paul Mackerras
2017-10-30 0:51 ` Ram Pai
2017-11-01 4:46 ` Michael Ellerman
2017-11-01 11:02 ` Paul Mackerras
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=adf2d270-29df-701f-206e-0d8a35084e47@linux.vnet.ibm.com \
--to=aneesh.kumar@linux.vnet.ibm.com \
--cc=benh@kernel.crashing.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mpe@ellerman.id.au \
--cc=paulus@ozlabs.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).