Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756703AbZKUSE1 (ORCPT ); Sat, 21 Nov 2009 13:04:27 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756666AbZKUSE0 (ORCPT ); Sat, 21 Nov 2009 13:04:26 -0500 Received: from mail-yx0-f187.google.com ([209.85.210.187]:33503 "EHLO mail-yx0-f187.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756648AbZKUSEZ (ORCPT ); Sat, 21 Nov 2009 13:04:25 -0500 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; b=UBeYndHMwMw4ad9dQl827fz5ltnav4izaQsDYDAxxH8QBrvCyYAcCegx4OQe3U0gmt frwEJsOtCC1iKAziVx185YQX179vB5POmnsL6PU71WJMbCLWZ+WB4qTu1+SyPDZ6DGxU JYc/TskRHyBL+qk5W5+3t7gbQVRb9qdVvpiAI= MIME-Version: 1.0 In-Reply-To: <4e5e476b0911201150p7cbdfc20k4b8d02f07e16abc8@mail.gmail.com> References: <20091116221827.GL13235@redhat.com> <4e5e476b0911180820y5d99a81et6be7f6f94442d0d5@mail.gmail.com> <20091118225626.GA2974@redhat.com> <4e5e476b0911181535y4d73d381s14b54c6d787d2b46@mail.gmail.com> <20091120141840.GA5872@redhat.com> <4e5e476b0911200628g42a0ab6ftd65b68bff5d1aea3@mail.gmail.com> <20091120150421.GC5872@redhat.com> <4e5e476b0911201032j5319b012v5fd0a64ebfe11089@mail.gmail.com> <20091120184257.GG5872@redhat.com> <4e5e476b0911201150p7cbdfc20k4b8d02f07e16abc8@mail.gmail.com> Date: Sat, 21 Nov 2009 18:57:47 +0100 Message-ID: <4e5e476b0911210957g5b099329u33fa6ec4d740f7d1@mail.gmail.com> Subject: Re: [RFC] Block IO Controller V2 - some results From: Corrado Zoccolo To: Vivek Goyal Cc: "Alan D. Brunelle" , linux-kernel@vger.kernel.org, jens.axboe@oracle.com Content-Type: multipart/mixed; boundary=001636c9283383643d0478e554aa Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4983 Lines: 86 --001636c9283383643d0478e554aa Content-Type: text/plain; charset=UTF-8 Hi Vivek, On Fri, Nov 20, 2009 at 8:50 PM, Corrado Zoccolo wrote: > On Fri, Nov 20, 2009 at 7:42 PM, Vivek Goyal wrote: >> >> Should we also take into account the "cfqq->dispatched" request in >> determining whether we should enable idling on deep queue random seeky >> readers? > > Probably, but I think the most important thing is to do some averaging > or hysteresis, so the completion of a single request doesn't cause the > switch of a workload. can you test the new version of the idling patch? I register the fact that the queue had a large depth in a flag, that is reset only when the idle times out (so at the end of the burst). Idling is enabled if that flag is set (and think time is acceptable). This should fix the switching behaviour you observed. I decided to not count cfqq->dispatched to determine the depth. In this way, when all queues in the system are random the idling is enabled only if the requests queue builds up faster than it can be consumed. Thanks, Corrado --001636c9283383643d0478e554aa Content-Type: application/octet-stream; name="0002-cfq-iosched-idling-on-deep-seeky-sync-queues.patch" Content-Disposition: attachment; filename="0002-cfq-iosched-idling-on-deep-seeky-sync-queues.patch" Content-Transfer-Encoding: base64 X-Attachment-Id: f_g2aocnxm0 RnJvbSAwMGZkMjIxYWIxNmE0ZWI1OTdkZmY1YjIzZjA2M2U1NDNmNWQ5OWI4IE1vbiBTZXAgMTcg MDA6MDA6MDAgMjAwMQpGcm9tOiBDb3JyYWRvIFpvY2NvbG8gPGNvcnJhZG9AZXQyLihub25lKT4K RGF0ZTogRnJpLCAyMCBOb3YgMjAwOSAxNDo1NTo1NCArMDEwMApTdWJqZWN0OiBbUEFUQ0ggMi8y XSBjZnEtaW9zY2hlZDogaWRsaW5nIG9uIGRlZXAgc2Vla3kgc3luYyBxdWV1ZXMKClNlZWt5IHN5 bmMgcXVldWVzIHdpdGggbGFyZ2UgZGVwdGggY2FuIGdhaW4gdW5mYWlybHkgYmlnIHNoYXJlIG9m IGRpc2sKdGltZSwgYXQgdGhlIGV4cGVuc2Ugb2Ygb3RoZXIgc2Vla3kgcXVldWVzLiBUaGlzIHBh dGNoIGVuc3VyZXMgdGhhdAppZGxpbmcgd2lsbCBiZSBlbmFibGVkIGZvciBxdWV1ZXMgd2l0aCBJ L08gZGVwdGggYXQgbGVhc3QgNCwgYW5kIHNtYWxsCnRoaW5rIHRpbWUuIFRoZSBkZWNpc2lvbiB0 byBlbmFibGUgaWRsaW5nIGlzIHN0aWNreSwgdW50aWwgYW4gaWRsZQp3aW5kb3cgdGltZXMgb3V0 IHdpdGhvdXQgc2VlaW5nIG5ldyByZXF1ZXN0LgoKVGhlIHJlYXNvbmluZyBiZWhpbmQgdGhlIGRl Y2lzaW9uIGlzIHRoYXQsIGlmIGFuIGFwcGxpY2F0aW9uIGlzIHVzaW5nCmxhcmdlIEkvTyBkZXB0 aCwgaXQgaXMgYWxyZWFkeSBvcHRpbWl6ZWQgdG8gbWFrZSBmdWxsIHV0aWxpemF0aW9uIG9mCnRo ZSBoYXJkd2FyZSwgYW5kIHRoZXJlZm9yZSB3ZSByZXNlcnZlIGEgc2xpY2Ugb2YgZXhjbHVzaXZl IHVzZSBmb3IgaXQuCgpTaWduZWQtb2ZmLWJ5OiBDb3JyYWRvIFpvY2NvbG8gPGN6b2Njb2xvQGdt YWlsLmNvbT4KLS0tCiBibG9jay9jZnEtaW9zY2hlZC5jIHwgICAxMyArKysrKysrKysrKystCiAx IGZpbGVzIGNoYW5nZWQsIDEyIGluc2VydGlvbnMoKyksIDEgZGVsZXRpb25zKC0pCgpkaWZmIC0t Z2l0IGEvYmxvY2svY2ZxLWlvc2NoZWQuYyBiL2Jsb2NrL2NmcS1pb3NjaGVkLmMKaW5kZXggMmEz MDRmNC4uMzczZTgwZiAxMDA2NDQKLS0tIGEvYmxvY2svY2ZxLWlvc2NoZWQuYworKysgYi9ibG9j ay9jZnEtaW9zY2hlZC5jCkBAIC0yNjAsNiArMjYwLDcgQEAgZW51bSBjZnFxX3N0YXRlX2ZsYWdz IHsKIAlDRlFfQ0ZRUV9GTEFHX3NsaWNlX25ldywJLyogbm8gcmVxdWVzdHMgZGlzcGF0Y2hlZCBp biBzbGljZSAqLwogCUNGUV9DRlFRX0ZMQUdfc3luYywJCS8qIHN5bmNocm9ub3VzIHF1ZXVlICov CiAJQ0ZRX0NGUVFfRkxBR19jb29wLAkJLyogY2ZxcSBpcyBzaGFyZWQgKi8KKwlDRlFfQ0ZRUV9G TEFHX2RlZXAsCQkvKiBzeW5jIGNmcXEgZXhwZXJpZW5jZWQgbGFyZ2UgZGVwdGggKi8KIH07CiAK ICNkZWZpbmUgQ0ZRX0NGUVFfRk5TKG5hbWUpCQkJCQkJXApAQCAtMjg2LDYgKzI4Nyw3IEBAIENG UV9DRlFRX0ZOUyhwcmlvX2NoYW5nZWQpOwogQ0ZRX0NGUVFfRk5TKHNsaWNlX25ldyk7CiBDRlFf Q0ZRUV9GTlMoc3luYyk7CiBDRlFfQ0ZRUV9GTlMoY29vcCk7CitDRlFfQ0ZRUV9GTlMoZGVlcCk7 CiAjdW5kZWYgQ0ZRX0NGUVFfRk5TCiAKICNkZWZpbmUgY2ZxX2xvZ19jZnFxKGNmcWQsIGNmcXEs IGZtdCwgYXJncy4uLikJXApAQCAtMjM1OSw4ICsyMzYxLDEyIEBAIGNmcV91cGRhdGVfaWRsZV93 aW5kb3coc3RydWN0IGNmcV9kYXRhICpjZnFkLCBzdHJ1Y3QgY2ZxX3F1ZXVlICpjZnFxLAogCiAJ ZW5hYmxlX2lkbGUgPSBvbGRfaWRsZSA9IGNmcV9jZnFxX2lkbGVfd2luZG93KGNmcXEpOwogCisJ aWYgKGNmcXEtPnF1ZXVlZFswXSArIGNmcXEtPnF1ZXVlZFsxXSA+PSA0KQorCQljZnFfbWFya19j ZnFxX2RlZXAoY2ZxcSk7CisKIAlpZiAoIWF0b21pY19yZWFkKCZjaWMtPmlvYy0+bnJfdGFza3Mp IHx8ICFjZnFkLT5jZnFfc2xpY2VfaWRsZSB8fAotCSAgICAoc2FtcGxlX3ZhbGlkKGNmcXEtPnNl ZWtfc2FtcGxlcykgJiYgQ0ZRUV9TRUVLWShjZnFxKSkpCisJICAgICghY2ZxX2NmcXFfZGVlcChj ZnFxKSAmJiBzYW1wbGVfdmFsaWQoY2ZxcS0+c2Vla19zYW1wbGVzKQorCSAgICAgJiYgQ0ZRUV9T RUVLWShjZnFxKSkpCiAJCWVuYWJsZV9pZGxlID0gMDsKIAllbHNlIGlmIChzYW1wbGVfdmFsaWQo Y2ljLT50dGltZV9zYW1wbGVzKSkgewogCQlpZiAoY2ljLT50dGltZV9tZWFuID4gY2ZxZC0+Y2Zx X3NsaWNlX2lkbGUpCkBAIC0yODU4LDYgKzI4NjQsMTEgQEAgc3RhdGljIHZvaWQgY2ZxX2lkbGVf c2xpY2VfdGltZXIodW5zaWduZWQgbG9uZyBkYXRhKQogCQkgKi8KIAkJaWYgKCFSQl9FTVBUWV9S T09UKCZjZnFxLT5zb3J0X2xpc3QpKQogCQkJZ290byBvdXRfa2ljazsKKworCQkvKgorCQkgKiBR dWV1ZSBkZXB0aCBmbGFnIGlzIHJlc2V0IG9ubHkgd2hlbiB0aGUgaWRsZSBkaWRuJ3Qgc3VjY2Vl ZAorCQkgKi8KKwkJY2ZxX2NsZWFyX2NmcXFfZGVlcChjZnFxKTsKIAl9CiBleHBpcmU6CiAJY2Zx X3NsaWNlX2V4cGlyZWQoY2ZxZCwgdGltZWRfb3V0KTsKLS0gCjEuNi4yLjUKCg== --001636c9283383643d0478e554aa-- -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/