From: Timofey Titovets <nefelim4ag@gmail.com> To: Srividya Desireddy <srividya.dr@samsung.com> Cc: "sjenning@redhat.com" <sjenning@redhat.com>, "ddstreet@ieee.org" <ddstreet@ieee.org>, "linux-mm@kvack.org" <linux-mm@kvack.org>, "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, "penberg@kernel.org" <penberg@kernel.org>, Dinakar Reddy Pathireddy <dinakar.p@samsung.com>, SHARAN ALLUR <sharan.allur@samsung.com>, RAJIB BASU <rajib.basu@samsung.com>, JUHUN KIM <juhunkim@samsung.com>, "srividya.desireddy@gmail.com" <srividya.desireddy@gmail.com> Subject: Re: [PATCH] zswap: Same-filled pages handling Date: Thu, 19 Oct 2017 00:31:18 +0300 [thread overview] Message-ID: <CAGqmi75Y9wbwBS0ZythcNF1gi6bW7g_XcuMDgLu=Nx4=pWC8Jw@mail.gmail.com> (raw) In-Reply-To: <20171018104832epcms5p1b2232e2236258de3d03d1344dde9fce0@epcms5p1> > +static int zswap_is_page_same_filled(void *ptr, unsigned long *value) > +{ > + unsigned int pos; > + unsigned long *page; > + > + page = (unsigned long *)ptr; > + for (pos = 1; pos < PAGE_SIZE / sizeof(*page); pos++) { > + if (page[pos] != page[0]) > + return 0; > + } > + *value = page[0]; > + return 1; > +} > + In theory you can speedup that check by memcmp(), And do something like first: memcmp(ptr, ptr + PAGE_SIZE/sizeof(*page)/2, PAGE_SIZE/2); After compare 1/4 with 2/4 Then 1/8 with 2/8. And after do you check with pattern, only on first 512 bytes. Just because memcmp() on fresh CPU are crazy fast. That can easy make you check less expensive. > +static void zswap_fill_page(void *ptr, unsigned long value) > +{ > + unsigned int pos; > + unsigned long *page; > + > + page = (unsigned long *)ptr; > + if (value == 0) > + memset(page, 0, PAGE_SIZE); > + else { > + for (pos = 0; pos < PAGE_SIZE / sizeof(*page); pos++) > + page[pos] = value; > + } > +} Same here, but with memcpy(). P.S. I'm just too busy to make fast performance test in user space, but my recent experience with that CPU commands, show what that make a sense: KSM patch: https://patchwork.kernel.org/patch/9980803/ User space tests: https://github.com/Nefelim4ag/memcmpe PAGE_SIZE: 65536, loop count: 1966080 memcmp: -28 time: 3216 ms, th: 40064.644611 MiB/s memcmpe: -28, offset: 62232 time: 3588 ms, th: 35902.462390 MiB/s memcmpe: -28, offset: 62232 time: 71 ms, th: 1792233.164286 MiB/s IIRC, with code like our, you must see ~2.5GiB/s Thanks. -- Have a nice day, Timofey.
WARNING: multiple messages have this Message-ID (diff)
From: Timofey Titovets <nefelim4ag@gmail.com> To: Srividya Desireddy <srividya.dr@samsung.com> Cc: "sjenning@redhat.com" <sjenning@redhat.com>, "ddstreet@ieee.org" <ddstreet@ieee.org>, "linux-mm@kvack.org" <linux-mm@kvack.org>, "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, "penberg@kernel.org" <penberg@kernel.org>, Dinakar Reddy Pathireddy <dinakar.p@samsung.com>, SHARAN ALLUR <sharan.allur@samsung.com>, RAJIB BASU <rajib.basu@samsung.com>, JUHUN KIM <juhunkim@samsung.com>, "srividya.desireddy@gmail.com" <srividya.desireddy@gmail.com> Subject: Re: [PATCH] zswap: Same-filled pages handling Date: Thu, 19 Oct 2017 00:31:18 +0300 [thread overview] Message-ID: <CAGqmi75Y9wbwBS0ZythcNF1gi6bW7g_XcuMDgLu=Nx4=pWC8Jw@mail.gmail.com> (raw) In-Reply-To: <20171018104832epcms5p1b2232e2236258de3d03d1344dde9fce0@epcms5p1> > +static int zswap_is_page_same_filled(void *ptr, unsigned long *value) > +{ > + unsigned int pos; > + unsigned long *page; > + > + page = (unsigned long *)ptr; > + for (pos = 1; pos < PAGE_SIZE / sizeof(*page); pos++) { > + if (page[pos] != page[0]) > + return 0; > + } > + *value = page[0]; > + return 1; > +} > + In theory you can speedup that check by memcmp(), And do something like first: memcmp(ptr, ptr + PAGE_SIZE/sizeof(*page)/2, PAGE_SIZE/2); After compare 1/4 with 2/4 Then 1/8 with 2/8. And after do you check with pattern, only on first 512 bytes. Just because memcmp() on fresh CPU are crazy fast. That can easy make you check less expensive. > +static void zswap_fill_page(void *ptr, unsigned long value) > +{ > + unsigned int pos; > + unsigned long *page; > + > + page = (unsigned long *)ptr; > + if (value == 0) > + memset(page, 0, PAGE_SIZE); > + else { > + for (pos = 0; pos < PAGE_SIZE / sizeof(*page); pos++) > + page[pos] = value; > + } > +} Same here, but with memcpy(). P.S. I'm just too busy to make fast performance test in user space, but my recent experience with that CPU commands, show what that make a sense: KSM patch: https://patchwork.kernel.org/patch/9980803/ User space tests: https://github.com/Nefelim4ag/memcmpe PAGE_SIZE: 65536, loop count: 1966080 memcmp: -28 time: 3216 ms, th: 40064.644611 MiB/s memcmpe: -28, offset: 62232 time: 3588 ms, th: 35902.462390 MiB/s memcmpe: -28, offset: 62232 time: 71 ms, th: 1792233.164286 MiB/s IIRC, with code like our, you must see ~2.5GiB/s Thanks. -- Have a nice day, Timofey. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2017-10-18 21:32 UTC|newest] Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top [not found] <CGME20171018104832epcms5p1b2232e2236258de3d03d1344dde9fce0@epcms5p1> 2017-10-18 10:48 ` [PATCH] zswap: Same-filled pages handling Srividya Desireddy 2017-10-18 10:48 ` Srividya Desireddy 2017-10-18 12:34 ` Matthew Wilcox 2017-10-18 12:34 ` Matthew Wilcox 2017-10-18 13:33 ` Timofey Titovets 2017-10-18 13:33 ` Timofey Titovets 2017-10-18 14:11 ` Matthew Wilcox 2017-10-18 14:11 ` Matthew Wilcox 2017-10-18 20:43 ` Andi Kleen 2017-10-18 20:43 ` Andi Kleen 2017-10-19 1:10 ` Matthew Wilcox 2017-10-19 1:10 ` Matthew Wilcox 2017-10-19 4:30 ` Andi Kleen 2017-10-19 4:30 ` Andi Kleen 2017-10-19 13:24 ` Matthew Wilcox 2017-10-19 13:24 ` Matthew Wilcox 2017-10-18 21:31 ` Timofey Titovets [this message] 2017-10-18 21:31 ` Timofey Titovets 2017-10-19 1:08 ` Matthew Wilcox 2017-10-19 1:08 ` Matthew Wilcox [not found] ` <CGME20171018104832epcms5p1b2232e2236258de3d03d1344dde9fce0@epcms5p3> 2017-11-02 15:08 ` Srividya Desireddy 2017-11-02 15:08 ` Srividya Desireddy 2017-11-17 22:10 ` Dan Streetman 2017-11-17 22:10 ` Dan Streetman 2017-11-17 22:07 ` Dan Streetman 2017-11-17 22:07 ` Dan Streetman 2017-11-17 21:27 ` Dan Streetman 2017-11-17 21:27 ` Dan Streetman 2017-11-20 23:46 ` Andrew Morton 2017-11-20 23:46 ` Andrew Morton 2017-11-28 11:35 ` Dan Streetman 2017-11-28 11:35 ` Dan Streetman [not found] ` <CGME20171018104832epcms5p1b2232e2236258de3d03d1344dde9fce0@epcms5p6> 2017-11-29 15:34 ` [PATCH] zswap: Update with same-value filled page feature Srividya Desireddy 2017-11-29 15:34 ` Srividya Desireddy 2017-11-29 21:29 ` Dan Streetman 2017-11-29 21:29 ` Dan Streetman 2017-12-06 11:48 ` [PATCH v2] " Srividya Desireddy 2017-12-06 11:48 ` Srividya Desireddy 2017-12-06 15:20 ` Dan Streetman 2017-12-06 15:20 ` Dan Streetman [not found] ` <CGME20171018104832epcms5p1b2232e2236258de3d03d1344dde9fce0@epcms5p4> 2017-11-21 14:18 ` [PATCH v2] zswap: Same-filled pages handling Srividya Desireddy 2017-11-21 14:18 ` Srividya Desireddy 2017-10-18 14:43 ` [PATCH] " Srividya Desireddy 2017-10-18 14:43 ` Srividya Desireddy
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='CAGqmi75Y9wbwBS0ZythcNF1gi6bW7g_XcuMDgLu=Nx4=pWC8Jw@mail.gmail.com' \ --to=nefelim4ag@gmail.com \ --cc=ddstreet@ieee.org \ --cc=dinakar.p@samsung.com \ --cc=juhunkim@samsung.com \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=penberg@kernel.org \ --cc=rajib.basu@samsung.com \ --cc=sharan.allur@samsung.com \ --cc=sjenning@redhat.com \ --cc=srividya.desireddy@gmail.com \ --cc=srividya.dr@samsung.com \ /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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.