From: Sunil Goutham <sunil.kovvuri@gmail.com> To: netdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Sunil.Goutham@caviumnetworks.com, Sunil Goutham <sgoutham@cavium.com> Subject: [PATCH 3/6] net: thunderx: Increase transmit queue length Date: Tue, 1 Dec 2015 14:43:40 +0530 [thread overview] Message-ID: <1448961223-41888-4-git-send-email-sunil.kovvuri@gmail.com> (raw) In-Reply-To: <y> From: Sunil Goutham <sgoutham@cavium.com> Under high transmit rates and with TSO enabled observing fluctuations in TX performance. Seen especially with iperf3 application. Since TSO is taken care at driver level, with 64KB of TSO packets and when window size is also high the rate at which CPU fills in transmit descriptors is much higher than what HW is able to process. Each 64KB TSO packet occupies gets segmented to ~43 1500 byte MTU packets and occupies ~130 descriptors. Hence increasing transmit queue size. Signed-off-by: Sunil Goutham <sgoutham@cavium.com> --- drivers/net/ethernet/cavium/thunder/nicvf_queues.h | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_queues.h b/drivers/net/ethernet/cavium/thunder/nicvf_queues.h index fb4957d..b1e93a9 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_queues.h +++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.h @@ -62,7 +62,7 @@ #define SND_QUEUE_CNT 8 #define CMP_QUEUE_CNT 8 /* Max of RCV and SND qcount */ -#define SND_QSIZE SND_QUEUE_SIZE2 +#define SND_QSIZE SND_QUEUE_SIZE3 #define SND_QUEUE_LEN (1ULL << (SND_QSIZE + 10)) #define MAX_SND_QUEUE_LEN (1ULL << (SND_QUEUE_SIZE6 + 10)) #define SND_QUEUE_THRESH 2ULL @@ -73,7 +73,7 @@ /* Keep CQ and SQ sizes same, if timestamping * is enabled this equation will change. */ -#define CMP_QSIZE CMP_QUEUE_SIZE2 +#define CMP_QSIZE CMP_QUEUE_SIZE3 #define CMP_QUEUE_LEN (1ULL << (CMP_QSIZE + 10)) #define CMP_QUEUE_CQE_THRESH 0 #define CMP_QUEUE_TIMER_THRESH 220 /* 10usec */ -- 1.7.1
WARNING: multiple messages have this Message-ID (diff)
From: sunil.kovvuri@gmail.com (Sunil Goutham) To: linux-arm-kernel@lists.infradead.org Subject: [PATCH 3/6] net: thunderx: Increase transmit queue length Date: Tue, 1 Dec 2015 14:43:40 +0530 [thread overview] Message-ID: <1448961223-41888-4-git-send-email-sunil.kovvuri@gmail.com> (raw) In-Reply-To: <y> From: Sunil Goutham <sgoutham@cavium.com> Under high transmit rates and with TSO enabled observing fluctuations in TX performance. Seen especially with iperf3 application. Since TSO is taken care at driver level, with 64KB of TSO packets and when window size is also high the rate at which CPU fills in transmit descriptors is much higher than what HW is able to process. Each 64KB TSO packet occupies gets segmented to ~43 1500 byte MTU packets and occupies ~130 descriptors. Hence increasing transmit queue size. Signed-off-by: Sunil Goutham <sgoutham@cavium.com> --- drivers/net/ethernet/cavium/thunder/nicvf_queues.h | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_queues.h b/drivers/net/ethernet/cavium/thunder/nicvf_queues.h index fb4957d..b1e93a9 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_queues.h +++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.h @@ -62,7 +62,7 @@ #define SND_QUEUE_CNT 8 #define CMP_QUEUE_CNT 8 /* Max of RCV and SND qcount */ -#define SND_QSIZE SND_QUEUE_SIZE2 +#define SND_QSIZE SND_QUEUE_SIZE3 #define SND_QUEUE_LEN (1ULL << (SND_QSIZE + 10)) #define MAX_SND_QUEUE_LEN (1ULL << (SND_QUEUE_SIZE6 + 10)) #define SND_QUEUE_THRESH 2ULL @@ -73,7 +73,7 @@ /* Keep CQ and SQ sizes same, if timestamping * is enabled this equation will change. */ -#define CMP_QSIZE CMP_QUEUE_SIZE2 +#define CMP_QSIZE CMP_QUEUE_SIZE3 #define CMP_QUEUE_LEN (1ULL << (CMP_QSIZE + 10)) #define CMP_QUEUE_CQE_THRESH 0 #define CMP_QUEUE_TIMER_THRESH 220 /* 10usec */ -- 1.7.1
next prev reply other threads:[~2015-12-01 9:12 UTC|newest] Thread overview: 102+ messages / expand[flat|nested] mbox.gz Atom feed top [not found] <y@samsung.com> 2012-05-22 5:57 ` [PATCH v3 2/2] regulator: Add support for MAX77686 yadi.brar01 2012-05-22 5:57 ` yadi.brar01 at gmail.com 2012-05-23 1:40 ` jonghwa3.lee 2012-05-23 1:40 ` jonghwa3.lee at samsung.com 2012-05-23 4:16 ` Yadwinder Singh Brar 2012-05-23 4:16 ` Yadwinder Singh Brar 2012-05-23 4:40 ` jonghwa3.lee 2012-05-23 4:40 ` jonghwa3.lee at samsung.com 2012-05-23 5:23 ` Yadwinder Singh Brar 2012-05-23 5:23 ` Yadwinder Singh Brar 2012-05-23 5:33 ` jonghwa3.lee 2012-05-23 5:33 ` jonghwa3.lee at samsung.com 2012-05-23 5:33 ` jonghwa3.lee 2012-05-23 10:18 ` Mark Brown 2012-05-23 10:18 ` Mark Brown 2012-05-23 13:02 ` Yadwinder Singh Brar 2012-05-23 13:02 ` Yadwinder Singh Brar 2012-05-23 6:08 ` Yadwinder Singh Brar 2012-05-23 6:08 ` Yadwinder Singh Brar 2012-05-23 1:50 ` jonghwa3.lee 2012-05-23 1:50 ` jonghwa3.lee at samsung.com 2012-05-23 4:17 ` Yadwinder Singh Brar 2012-05-23 4:17 ` Yadwinder Singh Brar 2014-07-23 1:40 ` [PATCH] extcon: Add missing REGMAP_I2C/REGMAP_IRQ dependency on extcon driver Chanwoo Choi 2014-07-23 8:20 ` Krzysztof Kozlowski 2014-07-25 8:39 ` Charles Keepax 2014-08-12 2:01 ` [PATCHv2 0/5] rtc: s3c: Refactoring s3c-rtc driver and support Exynos3250 RTC y 2014-08-12 2:01 ` [PATCHv2 1/5] rtc: s3c: Define s3c_rtc structure to remove global variables y 2014-08-22 20:42 ` Andrew Morton 2014-08-25 0:57 ` Chanwoo Choi 2014-08-26 21:31 ` Andrew Morton 2014-08-28 4:49 ` Chanwoo Choi 2014-08-12 2:01 ` [PATCHv2 2/5] rtc: s3c: Remove warning message when checking coding style with checkpatch script y 2014-08-12 2:01 ` [PATCHv2 3/5] rtc: s3c: Add s3c_rtc_data structure to use variant data instead of s3c_cpu_type y 2014-08-12 2:01 ` [PATCHv2 4/5] rtc: s3c: Add support for RTC of Exynos3250 SoC y 2014-08-12 2:01 ` [PATCHv2 5/5] ARM: dts: Fix wrong compatible string of Exynos3250 RTC dt node y 2015-12-01 9:13 ` Sunil Goutham [this message] 2015-12-01 9:13 ` [PATCH 3/6] net: thunderx: Increase transmit queue length Sunil Goutham 2015-12-01 14:40 ` Pavel Fedin 2015-12-01 14:40 ` Pavel Fedin 2015-12-01 15:33 ` Eric Dumazet 2015-12-01 15:33 ` Eric Dumazet 2015-12-01 16:30 ` Sunil Kovvuri 2015-12-01 16:30 ` Sunil Kovvuri 2015-12-01 19:30 ` David Miller 2015-12-01 19:30 ` David Miller 2015-12-02 5:48 ` Sunil Kovvuri 2015-12-02 5:48 ` Sunil Kovvuri 2015-12-02 13:25 ` Eric Dumazet 2015-12-02 13:25 ` Eric Dumazet 2015-12-02 16:50 ` Sunil Kovvuri 2015-12-02 16:50 ` Sunil Kovvuri 2015-12-02 16:59 ` Eric Dumazet 2015-12-02 16:59 ` Eric Dumazet 2015-12-02 17:31 ` David Miller 2015-12-02 17:31 ` David Miller 2015-12-02 9:05 ` Pavel Fedin 2015-12-02 9:05 ` Pavel Fedin 2015-12-02 10:31 ` Pavel Fedin 2015-12-02 10:31 ` Pavel Fedin 2015-12-02 12:29 ` Pavel Fedin 2015-12-02 12:29 ` Pavel Fedin 2015-12-02 12:57 ` Sunil Kovvuri 2015-12-02 12:57 ` Sunil Kovvuri 2015-12-02 13:22 ` Pavel Fedin 2015-12-02 13:22 ` Pavel Fedin 2015-12-02 8:09 ` Pavel Fedin 2015-12-02 8:09 ` Pavel Fedin 2015-12-01 9:13 ` [PATCH 5/6] net: thunderx: Switchon carrier only upon interface link up Sunil Goutham 2015-12-01 9:13 ` Sunil Goutham 2015-12-01 9:13 ` Sunil Goutham 2015-12-01 15:32 ` Pavel Fedin 2015-12-01 15:32 ` Pavel Fedin 2015-12-01 16:39 ` Sunil Kovvuri 2015-12-01 16:39 ` Sunil Kovvuri 2015-12-07 5:00 ` [PATCH 0/2] net: thunderx: Miscellaneous cleanups Sunil Goutham 2015-12-07 5:00 ` Sunil Goutham 2015-12-07 10:33 ` Pavel Fedin 2015-12-07 10:33 ` Pavel Fedin 2015-12-07 18:40 ` David Miller 2015-12-07 18:40 ` David Miller 2015-12-09 11:38 ` [PATCH 1/2] net: thunderx: HW TSO support for pass-2 hardware Sunil Goutham 2015-12-09 11:38 ` Sunil Goutham 2015-12-09 12:05 ` Pavel Fedin 2015-12-09 12:05 ` Pavel Fedin 2015-12-09 12:24 ` Sunil Kovvuri 2015-12-09 12:24 ` Sunil Kovvuri 2015-12-09 20:26 ` David Miller 2015-12-09 20:26 ` David Miller 2015-12-09 11:38 ` [PATCH 2/2] net: thunderx: Enable CQE count threshold interrupt Sunil Goutham 2015-12-09 11:38 ` Sunil Goutham 2015-12-09 12:07 ` Pavel Fedin 2015-12-09 12:07 ` Pavel Fedin 2015-12-09 12:26 ` Sunil Kovvuri 2015-12-09 12:26 ` Sunil Kovvuri 2015-12-10 7:55 ` [PATCH v2 0/2] net: thunderx: Support for pass-2 hw features Sunil Goutham 2015-12-10 7:55 ` Sunil Goutham 2015-12-10 7:55 ` Sunil Goutham 2015-12-10 8:52 ` Pavel Fedin 2015-12-10 8:52 ` Pavel Fedin 2015-12-12 4:38 ` David Miller 2015-12-12 4:38 ` David Miller
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=1448961223-41888-4-git-send-email-sunil.kovvuri@gmail.com \ --to=sunil.kovvuri@gmail.com \ --cc=Sunil.Goutham@caviumnetworks.com \ --cc=linux-arm-kernel@lists.infradead.org \ --cc=linux-kernel@vger.kernel.org \ --cc=netdev@vger.kernel.org \ --cc=sgoutham@cavium.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.