From: Stefan Agner <stefan@agner.ch>
To: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: dwmw2@infradead.org, computersforpeace@gmail.com,
mark.rutland@arm.com, boris.brezillon@free-electrons.com,
aaron@tastycactus.com, marb@ixxat.de, pawel.moll@arm.com,
ijc+devicetree@hellion.org.uk, linux-kernel@vger.kernel.org,
devicetree@vger.kernel.org, robh+dt@kernel.org,
linux-mtd@lists.infradead.org, kernel@pengutronix.de,
galak@codeaurora.org, shawn.guo@linaro.org,
linux-arm-kernel@lists.infradead.org, bpringlemeir@nbsps.com
Subject: Re: [PATCH v4 1/6] mtd: nand: vf610_nfc: Freescale NFC for VF610, MPC5125 and others
Date: Wed, 17 Jun 2015 15:02:03 +0200 [thread overview]
Message-ID: <32a7fe8d8df87dfe35ea50c582077c78@agner.ch> (raw)
In-Reply-To: <20150609200751.GC5720@breakpoint.cc>
On 2015-06-09 22:07, Sebastian Andrzej Siewior wrote:
> On 2015-06-03 17:05:16 [+0200], Stefan Agner wrote:
>> As Boris guessed correctly, the reason I used the raw variant was due to
>> performance improvements due to the barrier. However, I will use
>
> yeah, do you have any numbers by chance?
>
Just reevaluated the "performance optimizations". On a VF610 SoC (with
L2 cache) the improvements increase performance by 2.5% and below. On a
VF500 SoC (without L2 cache) it seems to have slightly more impact, up
to 4%. Overall, it seems to influence write more than read.
Back then, when I implemented the improvements it certainly had a bigger
impact. I don't have strong opinion on that...
=> VF610, with relaxed, optimized
[ 45.554246]
[ 45.555934] =================================================
[ 45.561762] mtd_speedtest: MTD device: 3
[ 45.576627] mtd_speedtest: MTD device size 104857600, eraseblock size
131072, page size 2048, count of eraseblocks 800, pages per eraseblock
64, OOB size 64
[ 45.595350] mtd_test: scanning for bad eraseblocks
[ 45.602177] mtd_test: scanned 800 eraseblocks, 0 are bad
[ 46.404682] mtd_speedtest: testing eraseblock write speed
[ 65.650704] mtd_speedtest: eraseblock write speed is 5344 KiB/s
[ 65.656738] mtd_speedtest: testing eraseblock read speed
[ 73.480687] mtd_speedtest: eraseblock read speed is 13118 KiB/s
[ 74.264593] mtd_speedtest: testing page write speed
[ 94.444190] mtd_speedtest: page write speed is 5087 KiB/s
[ 94.449674] mtd_speedtest: testing page read speed
[ 102.760387] mtd_speedtest: page read speed is 12343 KiB/s
[ 103.544838] mtd_speedtest: testing 2 page write speed
[ 123.257753] mtd_speedtest: 2 page write speed is 5199 KiB/s
[ 123.263440] mtd_speedtest: testing 2 page read speed
[ 131.171313] mtd_speedtest: 2 page read speed is 12963 KiB/s
[ 131.176981] mtd_speedtest: Testing erase speed
[ 131.968196] mtd_speedtest: erase speed is 130279 KiB/s
[ 131.973442] mtd_speedtest: Testing 2x multi-block erase speed
[ 132.153410] mtd_speedtest: 2x multi-block erase speed is 585142 KiB/s
[ 132.159924] mtd_speedtest: Testing 4x multi-block erase speed
[ 132.334864] mtd_speedtest: 4x multi-block erase speed is 609523 KiB/s
[ 132.341369] mtd_speedtest: Testing 8x multi-block erase speed
[ 132.516091] mtd_speedtest: 8x multi-block erase speed is 609523 KiB/s
[ 132.522602] mtd_speedtest: Testing 16x multi-block erase speed
[ 132.695800] mtd_speedtest: 16x multi-block erase speed is 613173
KiB/s
[ 132.702389] mtd_speedtest: Testing 32x multi-block erase speed
[ 132.874831] mtd_speedtest: 32x multi-block erase speed is 616867
KiB/s
[ 132.881445] mtd_speedtest: Testing 64x multi-block erase speed
[ 133.053647] mtd_speedtest: 64x multi-block erase speed is 616867
KiB/s
[ 133.060267] mtd_speedtest: finished
[ 133.063812] =================================================
=> VF610, without relaxed and using accessors
vf610_nfc_set/vf610_nfc_set_field and friends:
[ 60.015797]
[ 60.017481] =================================================
[ 60.023320] mtd_speedtest: MTD device: 3
[ 60.037232] mtd_speedtest: MTD device size 104857600, eraseblock size
131072, page size 2048, count of eraseblocks 800, pages per eraseblock
64, OOB size 64
[ 60.066359] mtd_test: scanning for bad eraseblocks
[ 60.074016] mtd_test: scanned 800 eraseblocks, 0 are bad
[ 60.286730] mtd_speedtest: testing eraseblock write speed
[ 79.679892] mtd_speedtest: eraseblock write speed is 5281 KiB/s
[ 79.685930] mtd_speedtest: testing eraseblock read speed
[ 87.563845] mtd_speedtest: eraseblock read speed is 13008 KiB/s
[ 88.353390] mtd_speedtest: testing page write speed
[ 108.984528] mtd_speedtest: page write speed is 4965 KiB/s
[ 108.990041] mtd_speedtest: testing page read speed
[ 117.012486] mtd_speedtest: page read speed is 12774 KiB/s
[ 117.801663] mtd_speedtest: testing 2 page write speed
[ 137.674009] mtd_speedtest: 2 page write speed is 5154 KiB/s
[ 137.679696] mtd_speedtest: testing 2 page read speed
[ 145.643303] mtd_speedtest: 2 page read speed is 12865 KiB/s
[ 145.648982] mtd_speedtest: Testing erase speed
[ 146.444423] mtd_speedtest: erase speed is 129456 KiB/s
[ 146.449669] mtd_speedtest: Testing 2x multi-block erase speed
[ 146.629634] mtd_speedtest: 2x multi-block erase speed is 588505 KiB/s
[ 146.636142] mtd_speedtest: Testing 4x multi-block erase speed
[ 146.813027] mtd_speedtest: 4x multi-block erase speed is 598830 KiB/s
[ 146.819577] mtd_speedtest: Testing 8x multi-block erase speed
[ 146.996654] mtd_speedtest: 8x multi-block erase speed is 595348 KiB/s
[ 147.003192] mtd_speedtest: Testing 16x multi-block erase speed
[ 147.178085] mtd_speedtest: 16x multi-block erase speed is 609523
KiB/s
[ 147.184703] mtd_speedtest: Testing 32x multi-block erase speed
[ 147.358306] mtd_speedtest: 32x multi-block erase speed is 613173
KiB/s
[ 147.364929] mtd_speedtest: Testing 64x multi-block erase speed
[ 147.540505] mtd_speedtest: 64x multi-block erase speed is 605917
KiB/s
[ 147.547106] mtd_speedtest: finished
[ 147.558336] =================================================
=> VF500, with relaxed, optimized
[ 42.878713]
[ 42.880775] =================================================
[ 42.886589] mtd_speedtest: MTD device: 3
[ 42.933491] mtd_speedtest: MTD device size 132120576, eraseblock size
131072, page size 2048, count of eraseblocks 1008, pages per eraseblock
64, OOB size 64
[ 42.981413] mtd_test: scanning for bad eraseblocks
[ 42.987505] mtd_test: block 142 is bad
[ 43.001342] mtd_test: block 1004 is bad
[ 43.024184] mtd_test: block 1005 is bad
[ 43.029184] mtd_test: block 1006 is bad
[ 43.045070] mtd_test: block 1007 is bad
[ 43.048975] mtd_test: scanned 1008 eraseblocks, 5 are bad
[ 43.854654] mtd_speedtest: testing eraseblock write speed
[ 68.873797] mtd_speedtest: eraseblock write speed is 5150 KiB/s
[ 68.879915] mtd_speedtest: testing eraseblock read speed
[ 80.409787] mtd_speedtest: eraseblock read speed is 11171 KiB/s
[ 81.053244] mtd_speedtest: testing page write speed
[ 109.851897] mtd_speedtest: page write speed is 4462 KiB/s
[ 109.857496] mtd_speedtest: testing page read speed
[ 121.964415] mtd_speedtest: page read speed is 10616 KiB/s
[ 122.603068] mtd_speedtest: testing 2 page write speed
[ 148.929905] mtd_speedtest: 2 page write speed is 4880 KiB/s
[ 148.935696] mtd_speedtest: testing 2 page read speed
[ 160.908554] mtd_speedtest: 2 page read speed is 10731 KiB/s
[ 160.914236] mtd_speedtest: Testing erase speed
[ 161.579429] mtd_speedtest: erase speed is 194226 KiB/s
[ 161.584680] mtd_speedtest: Testing 2x multi-block erase speed
[ 162.226599] mtd_speedtest: 2x multi-block erase speed is 201861 KiB/s
[ 162.233153] mtd_speedtest: Testing 4x multi-block erase speed
[ 162.870923] mtd_speedtest: 4x multi-block erase speed is 202818 KiB/s
[ 162.877591] mtd_speedtest: Testing 8x multi-block erase speed
[ 163.503176] mtd_speedtest: 8x multi-block erase speed is 207070 KiB/s
[ 163.509826] mtd_speedtest: Testing 16x multi-block erase speed
[ 164.138392] mtd_speedtest: 16x multi-block erase speed is 206405
KiB/s
[ 164.145128] mtd_speedtest: Testing 32x multi-block erase speed
[ 164.768705] mtd_speedtest: 32x multi-block erase speed is 207741
KiB/s
[ 164.775475] mtd_speedtest: Testing 64x multi-block erase speed
[ 165.394716] mtd_speedtest: 64x multi-block erase speed is 209094
KiB/s
[ 165.401347] mtd_speedtest: finished
[ 165.415581] =================================================
=> VF500, without relaxed and using accessors
[ 93.466642]
[ 93.468578] =================================================
[ 93.474563] mtd_speedtest: MTD device: 3
[ 93.510699] mtd_speedtest: MTD device size 132120576, eraseblock size
131072, page size 2048, count of eraseblocks 1008, pages per eraseblock
64, OOB size 64
[ 93.534490] mtd_test: scanning for bad eraseblocks
[ 93.553066] mtd_test: block 142 is bad
[ 93.586518] mtd_test: block 1004 is bad
[ 93.590559] mtd_test: block 1005 is bad
[ 93.595654] mtd_test: block 1006 is bad
[ 93.607090] mtd_test: block 1007 is bad
[ 93.630667] mtd_test: scanned 1008 eraseblocks, 5 are bad
[ 94.457510] mtd_speedtest: testing eraseblock write speed
[ 120.495695] mtd_speedtest: eraseblock write speed is 4932 KiB/s
[ 120.501726] mtd_speedtest: testing eraseblock read speed
[ 131.937006] mtd_speedtest: eraseblock read speed is 11233 KiB/s
[ 132.576468] mtd_speedtest: testing page write speed
[ 160.663012] mtd_speedtest: page write speed is 4572 KiB/s
[ 160.668608] mtd_speedtest: testing page read speed
[ 172.624020] mtd_speedtest: page read speed is 10743 KiB/s
[ 173.263221] mtd_speedtest: testing 2 page write speed
[ 199.811255] mtd_speedtest: 2 page write speed is 4837 KiB/s
[ 199.816933] mtd_speedtest: testing 2 page read speed
[ 211.558975] mtd_speedtest: 2 page read speed is 10939 KiB/s
[ 211.564657] mtd_speedtest: Testing erase speed
[ 212.227608] mtd_speedtest: erase speed is 194816 KiB/s
[ 212.232951] mtd_speedtest: Testing 2x multi-block erase speed
[ 212.875607] mtd_speedtest: 2x multi-block erase speed is 201544 KiB/s
[ 212.882254] mtd_speedtest: Testing 4x multi-block erase speed
[ 213.521870] mtd_speedtest: 4x multi-block erase speed is 202818 KiB/s
[ 213.528471] mtd_speedtest: Testing 8x multi-block erase speed
[ 214.154043] mtd_speedtest: 8x multi-block erase speed is 207070 KiB/s
[ 214.160690] mtd_speedtest: Testing 16x multi-block erase speed
[ 214.785910] mtd_speedtest: 16x multi-block erase speed is 207405
KiB/s
[ 214.792652] mtd_speedtest: Testing 32x multi-block erase speed
[ 215.418150] mtd_speedtest: 32x multi-block erase speed is 207405
KiB/s
[ 215.424883] mtd_speedtest: Testing 64x multi-block erase speed
[ 216.051367] mtd_speedtest: 64x multi-block erase speed is 206737
KiB/s
[ 216.058056] mtd_speedtest: finished
[ 216.069363] =================================================
--
Stefan
next prev parent reply other threads:[~2015-06-17 13:05 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-25 16:28 [PATCH v4 0/6] mtd: nand: vf610_nfc: Freescale NFC for VF610 Stefan Agner
2015-03-25 16:28 ` [PATCH v4 1/6] mtd: nand: vf610_nfc: Freescale NFC for VF610, MPC5125 and others Stefan Agner
2015-06-02 1:32 ` Brian Norris
2015-06-02 9:01 ` Stefan Agner
2015-06-03 8:10 ` Sebastian Andrzej Siewior
2015-06-03 9:45 ` Stefan Agner
2015-06-03 12:03 ` Sebastian Andrzej Siewior
2015-06-03 15:09 ` Stefan Agner
2015-06-09 20:05 ` Sebastian Andrzej Siewior
2015-06-03 13:08 ` Sebastian Andrzej Siewior
2015-06-03 15:05 ` Stefan Agner
2015-06-09 20:07 ` Sebastian Andrzej Siewior
2015-06-09 20:31 ` Stefan Agner
2015-06-12 19:44 ` Sebastian Andrzej Siewior
2015-06-17 13:02 ` Stefan Agner [this message]
2015-06-18 18:27 ` Sebastian Andrzej Siewior
2015-03-25 16:28 ` [PATCH v4 2/6] mtd: nand: vf610_nfc: add hardware BCH-ECC support Stefan Agner
2015-03-25 16:28 ` [PATCH v4 3/6] mtd: nand: vf610_nfc: add device tree bindings Stefan Agner
2015-03-25 16:28 ` [PATCH v4 4/6] ARM: vf610: enable NAND Flash Controller Stefan Agner
2015-03-25 16:28 ` [PATCH v4 5/6] ARM: dts: vf610: add NAND flash controller peripherial Stefan Agner
2015-03-25 16:28 ` [PATCH v4 6/6] ARM: dts: vf-colibri: enable NAND flash controller Stefan Agner
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=32a7fe8d8df87dfe35ea50c582077c78@agner.ch \
--to=stefan@agner.ch \
--cc=aaron@tastycactus.com \
--cc=boris.brezillon@free-electrons.com \
--cc=bpringlemeir@nbsps.com \
--cc=computersforpeace@gmail.com \
--cc=devicetree@vger.kernel.org \
--cc=dwmw2@infradead.org \
--cc=galak@codeaurora.org \
--cc=ijc+devicetree@hellion.org.uk \
--cc=kernel@pengutronix.de \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mtd@lists.infradead.org \
--cc=marb@ixxat.de \
--cc=mark.rutland@arm.com \
--cc=pawel.moll@arm.com \
--cc=robh+dt@kernel.org \
--cc=sebastian@breakpoint.cc \
--cc=shawn.guo@linaro.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).