RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Dear All, Our tests show that CPU usage is always 100% during netperf test. So the speed of CPU is important in the overall performance of the gigabit link. If we can increase the CPU core clock frequency we may achieve better results using existing hardware/software configuration. I know that the PPC core inside FX12 can run with clock frequencies up to 450MHz, However Base System Builder for ML403 allows just frequencies of up to 300MHz for PPC core. Does any body here know how I can make PPC core to run with 400MHz in ML403? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Tue, 26 Jun 2007 18:12:55 + Dear Mohammad, ML403---PC : 410Mbits/s PC---ML403 : 210Mbits/s These results are interesting. In priciple, board to PC will be less than PC to board. And also you board to PC speed is quite fast. I never had that high before. :) We have described the characteristics of our base system in previous posts here In additiona we have : 1- enabled the ppc caches This will help the improvement quite a lot. 2- we have set BD_IN_BRAM in adapter.c to 1. ( default is 0 ) Actually I didn't try to modify this before. My previous results are based on bd_NOT_in_bram. :) From my understanding, enable this option will put the buffer descriptor in BRAM rather than DDR. Perhaps I can also try it and to see if there is any improvement on my system. TX_THRESHOLD is 16 and RX_THRESHOLD is 2. the virtex4 fx12 device on ML403 is now completely full, we do not have any free block memories nor any logic slices. Maybe if we had more space we could choose higher values for XTE_SEND_BD_CNT and XTE_RECV_BD_CNT i.e. 384. Do you think this will improve performance? Probably yes. But I never modified these numbers before. My default ones are 512 respectively. There is also another interesting test, We executed netperf on both of PC and ML403 simultanously , when we do not put BDs in BRAM, the performance of ML403--PC link drops from 390Mbits to 45Mbits, but when using PLB BRAMs for BDs the performance decreases from 410Mbits/s to just 130Mbita/s. It is important when the user wants to transfer data in both directions simulatanously. Definitely! The bottleneck is CPU processing capability. So if you send and receive data at the same time, the results will be much worse. I think another reason is TCP is guaranteed protocal. So there will be some acknowledgements returning back when you send packets out. Thus your bandwidth will be taken a little away. However compared with the consumption of CPU, this perhaps will be trivial. BR Ming From: [EMAIL PROTECTED] To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Mon, 25 Jun 2007 10:03:30 + Dear Mohammad, The results are as follows: PC--ML403 TCP_SENDFILE : 38Mbps ML403---PC TCP_SENDFILE: 155Mbps This result is unreasonable. Because PC is more powerful than your board, so PC-board should be faster than board-PC. The transfer rate from ML403 to PC has improved by a factor of 2, I see on the posts here in the mailing list that you have reached a band width of 301Mbps. Yes, with all features which could improve performance enabled, we can get around 300Mbps for TCP transfer. one more hint, did you enable caches on your system? perhaps it will help. Anyway, double check your hardware design to make sure all features are enabled.That's all I can suggest. BR Ming From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:10:16 + Use the following command in Linux please: ifconfig eth0 mtu 8982 As well you should do that on your PC in the measurement. Ming From: Mohammad Sadegh Sadri To: Ming Liu , ,, Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:08:29 + Dear Ming, Really thanks for reply, about thresholds and waitbound OK! I'll adjust them in adapter.c , but what about enabling jumbo frames? should I do any thing special to enable Jumbo fram support? we were thinking that it is enabled by default. Is it? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Actually I have asked the xilinx expert on the statistics. With the PLB_TEMAC, we can also get a result like that, say 300Mbps for TCP. (From their numbers, the throughput is even higher.) Some remindings from my experience: Remember to enable everything in the hardware and software which can improve the performance, such as, CS_offloading, data allighment engines, large fifos, and so on, from the configuration of PLB_TEMAC in EDK. As well, remember to enable cache. In the software field, interrupt_coaleascing will also help. At this time, normally we can get more than 100Mbps for TCP. Jumbo-frame of 8982 will almost double this number. Have fun. BR Ming From: Greg Crocker [EMAIL PROTECTED] To: linuxppc-embedded@ozlabs.org Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Mon, 25 Jun 2007 15:42:09 -0400 I was able to achieve ~320 Mbit/sec data rate using the Gigabit System Reference Design (GSRD XAPP535/536) from Xilinx. This utilizes the LocalLink TEMAC to perform the transfers. The reference design provides the Linux 2.4 drivers that can be ported to Linux 2.6 with a little effort. This implementation did not use checksum offloading and the data rates were achieved using TCP_STREAM on netperf. Greg ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded _ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Dear Ming, Thanks to your comments , Our tests now give the following results: ML403---PC : 410Mbits/s PC---ML403 : 210Mbits/s We have described the characteristics of our base system in previous posts here In additiona we have : 1- enabled the ppc caches 2- we have set BD_IN_BRAM in adapter.c to 1. ( default is 0 ) TX_THRESHOLD is 16 and RX_THRESHOLD is 2. the virtex4 fx12 device on ML403 is now completely full, we do not have any free block memories nor any logic slices. Maybe if we had more space we could choose higher values for XTE_SEND_BD_CNT and XTE_RECV_BD_CNT i.e. 384. Do you think this will improve performance? There is also another interesting test, We executed netperf on both of PC and ML403 simultanously , when we do not put BDs in BRAM, the performance of ML403--PC link drops from 390Mbits to 45Mbits, but when using PLB BRAMs for BDs the performance decreases from 410Mbits/s to just 130Mbita/s. It is important when the user wants to transfer data in both directions simulatanously. Thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Mon, 25 Jun 2007 10:03:30 + Dear Mohammad, The results are as follows: PC--ML403 TCP_SENDFILE : 38Mbps ML403---PC TCP_SENDFILE: 155Mbps This result is unreasonable. Because PC is more powerful than your board, so PC-board should be faster than board-PC. The transfer rate from ML403 to PC has improved by a factor of 2, I see on the posts here in the mailing list that you have reached a band width of 301Mbps. Yes, with all features which could improve performance enabled, we can get around 300Mbps for TCP transfer. one more hint, did you enable caches on your system? perhaps it will help. Anyway, double check your hardware design to make sure all features are enabled.That's all I can suggest. BR Ming From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:10:16 + Use the following command in Linux please: ifconfig eth0 mtu 8982 As well you should do that on your PC in the measurement. Ming From: Mohammad Sadegh Sadri To: Ming Liu , ,, Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:08:29 + Dear Ming, Really thanks for reply, about thresholds and waitbound OK! I'll adjust them in adapter.c , but what about enabling jumbo frames? should I do any thing special to enable Jumbo fram support? we were thinking that it is enabled by default. Is it? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 18:48:19 + Dear Mohammad, There are some parameters which could be adjusted to improve the performance. They are: TX and RX_Threshold TX and RX_waitbound. In my system, we use TX_Threshold=16 and Rx_Threshold=8 and both waitbound=1. Also Jumbo frame of 8982 could be enable. Try those hints and share your improvement with us. BR Ming From: Mohammad Sadegh Sadri To: Andrei Konovalov , Linux PPC Linux PPC, Grant Likely Subject: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 12:19:12 + Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all of the posts here carefully: these are the system characteristics; Board : ML403 EDK: EDK9.1SP2 Hard TEMAC version and PLTEMAC version are both 3.0.a PPC clock frequency : 300MHz Kernel : 2.6.21-rc7 , downloaded from grant's git tree some thing near one week ago DMA type: 3 (sg dma) DRE : enabled for TX and RX, (2) CSUM offload is enabled for both of TX and RX tx and rx fifo sizes : 131072 bits the board comes up over NFS root file system completely and without any problems. PC system used for these tests is : CPU P4 Dual Core, 3.4GHz , 2Gigabytes memory, Dual gigabit ethernet port, running linux 2.6.21.3 We have tested the PC system band width and it can easily reach 966mbits/s when connected to the same PC. ( using the same cross cable used for ml403 test) Netperf is compiled with TCP SEND FILE enabled, ( -DHAVE_SENDFILE) (from board to PC) netperf -t TCP_SENDFILE
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Dear Mohammad, ML403---PC : 410Mbits/s PC---ML403 : 210Mbits/s These results are interesting. In priciple, board to PC will be less than PC to board. And also you board to PC speed is quite fast. I never had that high before. :) We have described the characteristics of our base system in previous posts here In additiona we have : 1- enabled the ppc caches This will help the improvement quite a lot. 2- we have set BD_IN_BRAM in adapter.c to 1. ( default is 0 ) Actually I didn't try to modify this before. My previous results are based on bd_NOT_in_bram. :) From my understanding, enable this option will put the buffer descriptor in BRAM rather than DDR. Perhaps I can also try it and to see if there is any improvement on my system. TX_THRESHOLD is 16 and RX_THRESHOLD is 2. the virtex4 fx12 device on ML403 is now completely full, we do not have any free block memories nor any logic slices. Maybe if we had more space we could choose higher values for XTE_SEND_BD_CNT and XTE_RECV_BD_CNT i.e. 384. Do you think this will improve performance? Probably yes. But I never modified these numbers before. My default ones are 512 respectively. There is also another interesting test, We executed netperf on both of PC and ML403 simultanously , when we do not put BDs in BRAM, the performance of ML403--PC link drops from 390Mbits to 45Mbits, but when using PLB BRAMs for BDs the performance decreases from 410Mbits/s to just 130Mbita/s. It is important when the user wants to transfer data in both directions simulatanously. Definitely! The bottleneck is CPU processing capability. So if you send and receive data at the same time, the results will be much worse. I think another reason is TCP is guaranteed protocal. So there will be some acknowledgements returning back when you send packets out. Thus your bandwidth will be taken a little away. However compared with the consumption of CPU, this perhaps will be trivial. BR Ming From: [EMAIL PROTECTED] To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Mon, 25 Jun 2007 10:03:30 + Dear Mohammad, The results are as follows: PC--ML403 TCP_SENDFILE : 38Mbps ML403---PC TCP_SENDFILE: 155Mbps This result is unreasonable. Because PC is more powerful than your board, so PC-board should be faster than board-PC. The transfer rate from ML403 to PC has improved by a factor of 2, I see on the posts here in the mailing list that you have reached a band width of 301Mbps. Yes, with all features which could improve performance enabled, we can get around 300Mbps for TCP transfer. one more hint, did you enable caches on your system? perhaps it will help. Anyway, double check your hardware design to make sure all features are enabled.That's all I can suggest. BR Ming From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:10:16 + Use the following command in Linux please: ifconfig eth0 mtu 8982 As well you should do that on your PC in the measurement. Ming From: Mohammad Sadegh Sadri To: Ming Liu , ,, Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:08:29 + Dear Ming, Really thanks for reply, about thresholds and waitbound OK! I'll adjust them in adapter.c , but what about enabling jumbo frames? should I do any thing special to enable Jumbo fram support? we were thinking that it is enabled by default. Is it? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 18:48:19 + Dear Mohammad, There are some parameters which could be adjusted to improve the performance. They are: TX and RX_Threshold TX and RX_waitbound. In my system, we use TX_Threshold=16 and Rx_Threshold=8 and both waitbound=1. Also Jumbo frame of 8982 could be enable. Try those hints and share your improvement with us. BR Ming From: Mohammad Sadegh Sadri To: Andrei Konovalov , Linux PPC Linux PPC, Grant Likely Subject: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 12:19:12 + Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Dear Mohammad, The results are as follows: PC--ML403 TCP_SENDFILE : 38Mbps ML403---PC TCP_SENDFILE: 155Mbps This result is unreasonable. Because PC is more powerful than your board, so PC-board should be faster than board-PC. The transfer rate from ML403 to PC has improved by a factor of 2, I see on the posts here in the mailing list that you have reached a band width of 301Mbps. Yes, with all features which could improve performance enabled, we can get around 300Mbps for TCP transfer. one more hint, did you enable caches on your system? perhaps it will help. Anyway, double check your hardware design to make sure all features are enabled.That's all I can suggest. BR Ming From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:10:16 + Use the following command in Linux please: ifconfig eth0 mtu 8982 As well you should do that on your PC in the measurement. Ming From: Mohammad Sadegh Sadri To: Ming Liu , ,, Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:08:29 + Dear Ming, Really thanks for reply, about thresholds and waitbound OK! I'll adjust them in adapter.c , but what about enabling jumbo frames? should I do any thing special to enable Jumbo fram support? we were thinking that it is enabled by default. Is it? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 18:48:19 + Dear Mohammad, There are some parameters which could be adjusted to improve the performance. They are: TX and RX_Threshold TX and RX_waitbound. In my system, we use TX_Threshold=16 and Rx_Threshold=8 and both waitbound=1. Also Jumbo frame of 8982 could be enable. Try those hints and share your improvement with us. BR Ming From: Mohammad Sadegh Sadri To: Andrei Konovalov , Linux PPC Linux PPC, Grant Likely Subject: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 12:19:12 + Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all of the posts here carefully: these are the system characteristics; Board : ML403 EDK: EDK9.1SP2 Hard TEMAC version and PLTEMAC version are both 3.0.a PPC clock frequency : 300MHz Kernel : 2.6.21-rc7 , downloaded from grant's git tree some thing near one week ago DMA type: 3 (sg dma) DRE : enabled for TX and RX, (2) CSUM offload is enabled for both of TX and RX tx and rx fifo sizes : 131072 bits the board comes up over NFS root file system completely and without any problems. PC system used for these tests is : CPU P4 Dual Core, 3.4GHz , 2Gigabytes memory, Dual gigabit ethernet port, running linux 2.6.21.3 We have tested the PC system band width and it can easily reach 966mbits/s when connected to the same PC. ( using the same cross cable used for ml403 test) Netperf is compiled with TCP SEND FILE enabled, ( -DHAVE_SENDFILE) (from board to PC) netperf -t TCP_SENDFILE -H 10.10.10.250 -F /boot/zImage.elf -- -m 16384 -s 87380 -S 87380 the measured bandwidth for this test was just 40.66Mbits. It is also true for netperf from PC to board. we do not have any more idea about what we should do to improve the bandwidth. any help or ideas is appreciated... _ Connect to the next generation of MSN Messenger?http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-ussource=wlmailtagline ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ _ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ _ Discover the new Windows Vista http://search.msn.com/results.aspx?q=windows+vistamkt=en-USform=QBRE _ 免费下载 MSN
Re: ML403 gigabit ethernet bandwidth - 2.6 kernel
Hi, We need to findout where is the bottlenect. 1. Run vmstat on the ML403 board and find out the percentage CPU is busy when you are transferring the file. That will show if cpu is busy or not. 2. Run oprofile and find out which are the routines eating away the cpu time. Once we have data from both the above routines, we can find out the bottlenecks. Regards Bhupi On 6/23/07, Mohammad Sadegh Sadri [EMAIL PROTECTED] wrote: Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all of the posts here carefully: these are the system characteristics; Board : ML403 EDK: EDK9.1SP2 Hard TEMAC version and PLTEMAC version are both 3.0.a PPC clock frequency : 300MHz Kernel : 2.6.21-rc7 , downloaded from grant's git tree some thing near one week ago DMA type: 3 (sg dma) DRE : enabled for TX and RX, (2) CSUM offload is enabled for both of TX and RX tx and rx fifo sizes : 131072 bits the board comes up over NFS root file system completely and without any problems. PC system used for these tests is : CPU P4 Dual Core, 3.4GHz , 2Gigabytes memory, Dual gigabit ethernet port, running linux 2.6.21.3 We have tested the PC system band width and it can easily reach 966mbits/s when connected to the same PC. ( using the same cross cable used for ml403 test) Netperf is compiled with TCP SEND FILE enabled, ( -DHAVE_SENDFILE) (from board to PC) netperf -t TCP_SENDFILE -H 10.10.10.250 -F /boot/zImage.elf -- -m 16384 -s 87380 -S 87380 the measured bandwidth for this test was just 40.66Mbits. It is also true for netperf from PC to board. we do not have any more idea about what we should do to improve the bandwidth. any help or ideas is appreciated... _ Connect to the next generation of MSN Messenger http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-ussource=wlmailtagline ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
I was able to achieve ~320 Mbit/sec data rate using the Gigabit System Reference Design (GSRD XAPP535/536) from Xilinx. This utilizes the LocalLink TEMAC to perform the transfers. The reference design provides the Linux 2.4 drivers that can be ported to Linux 2.6 with a little effort. This implementation did not use checksum offloading and the data rates were achieved using TCP_STREAM on netperf. Greg ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Dear Ming We have changed our system characteristics to have TX_THRESHOLD=16 and RX_THRESHOLD=8 and in addition we enabled jumbo frames of 8982 bytes. The results are as follows: PC--ML403 TCP_SENDFILE : 38Mbps ML403---PC TCP_SENDFILE: 155Mbps The transfer rate from ML403 to PC has improved by a factor of 2, I see on the posts here in the mailing list that you have reached a band width of 301Mbps. we are also wondering why we do not have any improve in PC to ML403 bandwidth also we observed if TX_THRESHOLD=16 and RX_THRESHOLD=2, then PC to ML403 bandwidth will increase to some thing near 60Mbps. From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:10:16 + Use the following command in Linux please: ifconfig eth0 mtu 8982 As well you should do that on your PC in the measurement. Ming From: Mohammad Sadegh Sadri To: Ming Liu , ,, Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:08:29 + Dear Ming, Really thanks for reply, about thresholds and waitbound OK! I'll adjust them in adapter.c , but what about enabling jumbo frames? should I do any thing special to enable Jumbo fram support? we were thinking that it is enabled by default. Is it? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 18:48:19 + Dear Mohammad, There are some parameters which could be adjusted to improve the performance. They are: TX and RX_Threshold TX and RX_waitbound. In my system, we use TX_Threshold=16 and Rx_Threshold=8 and both waitbound=1. Also Jumbo frame of 8982 could be enable. Try those hints and share your improvement with us. BR Ming From: Mohammad Sadegh Sadri To: Andrei Konovalov , Linux PPC Linux PPC, Grant Likely Subject: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 12:19:12 + Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all of the posts here carefully: these are the system characteristics; Board : ML403 EDK: EDK9.1SP2 Hard TEMAC version and PLTEMAC version are both 3.0.a PPC clock frequency : 300MHz Kernel : 2.6.21-rc7 , downloaded from grant's git tree some thing near one week ago DMA type: 3 (sg dma) DRE : enabled for TX and RX, (2) CSUM offload is enabled for both of TX and RX tx and rx fifo sizes : 131072 bits the board comes up over NFS root file system completely and without any problems. PC system used for these tests is : CPU P4 Dual Core, 3.4GHz , 2Gigabytes memory, Dual gigabit ethernet port, running linux 2.6.21.3 We have tested the PC system band width and it can easily reach 966mbits/s when connected to the same PC. ( using the same cross cable used for ml403 test) Netperf is compiled with TCP SEND FILE enabled, ( -DHAVE_SENDFILE) (from board to PC) netperf -t TCP_SENDFILE -H 10.10.10.250 -F /boot/zImage.elf -- -m 16384 -s 87380 -S 87380 the measured bandwidth for this test was just 40.66Mbits. It is also true for netperf from PC to board. we do not have any more idea about what we should do to improve the bandwidth. any help or ideas is appreciated... _ Connect to the next generation of MSN Messenger?http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-ussource=wlmailtagline ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ _ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ _ Discover the new Windows Vista http://search.msn.com/results.aspx?q=windows+vistamkt=en-USform=QBRE ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Dear Mohammad, There are some parameters which could be adjusted to improve the performance. They are: TX and RX_Threshold TX and RX_waitbound. In my system, we use TX_Threshold=16 and Rx_Threshold=8 and both waitbound=1. Also Jumbo frame of 8982 could be enable. Try those hints and share your improvement with us. BR Ming From: Mohammad Sadegh Sadri [EMAIL PROTECTED] To: Andrei Konovalov [EMAIL PROTECTED], Linux PPC Linux PPClinuxppc-embedded@ozlabs.org, Grant Likely [EMAIL PROTECTED] Subject: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 12:19:12 + Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all of the posts here carefully: these are the system characteristics; Board : ML403 EDK: EDK9.1SP2 Hard TEMAC version and PLTEMAC version are both 3.0.a PPC clock frequency : 300MHz Kernel : 2.6.21-rc7 , downloaded from grant's git tree some thing near one week ago DMA type: 3 (sg dma) DRE : enabled for TX and RX, (2) CSUM offload is enabled for both of TX and RX tx and rx fifo sizes : 131072 bits the board comes up over NFS root file system completely and without any problems. PC system used for these tests is : CPU P4 Dual Core, 3.4GHz , 2Gigabytes memory, Dual gigabit ethernet port, running linux 2.6.21.3 We have tested the PC system band width and it can easily reach 966mbits/s when connected to the same PC. ( using the same cross cable used for ml403 test) Netperf is compiled with TCP SEND FILE enabled, ( -DHAVE_SENDFILE) (from board to PC) netperf -t TCP_SENDFILE -H 10.10.10.250 -F /boot/zImage.elf -- -m 16384 -s 87380 -S 87380 the measured bandwidth for this test was just 40.66Mbits. It is also true for netperf from PC to board. we do not have any more idea about what we should do to improve the bandwidth. any help or ideas is appreciated... _ Connect to the next generation of MSN Messenger?http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-ussource=wlmailtagline ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded
RE: ML403 gigabit ethernet bandwidth - 2.6 kernel
Use the following command in Linux please: ifconfig eth0 mtu 8982 As well you should do that on your PC in the measurement. Ming From: Mohammad Sadegh Sadri [EMAIL PROTECTED] To: Ming Liu [EMAIL PROTECTED], [EMAIL PROTECTED],linuxppc-embedded@ozlabs.org, [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 19:08:29 + Dear Ming, Really thanks for reply, about thresholds and waitbound OK! I'll adjust them in adapter.c , but what about enabling jumbo frames? should I do any thing special to enable Jumbo fram support? we were thinking that it is enabled by default. Is it? thanks From: [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED]; linuxppc-embedded@ozlabs.org; [EMAIL PROTECTED] Subject: RE: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 18:48:19 + Dear Mohammad, There are some parameters which could be adjusted to improve the performance. They are: TX and RX_Threshold TX and RX_waitbound. In my system, we use TX_Threshold=16 and Rx_Threshold=8 and both waitbound=1. Also Jumbo frame of 8982 could be enable. Try those hints and share your improvement with us. BR Ming From: Mohammad Sadegh Sadri To: Andrei Konovalov , Linux PPC Linux PPC, Grant Likely Subject: ML403 gigabit ethernet bandwidth - 2.6 kernel Date: Sat, 23 Jun 2007 12:19:12 + Dear all, Recently we did a set of tests on performance of virtex 4FX hard TEMAC module using ML403 we studied all of the posts here carefully: these are the system characteristics; Board : ML403 EDK: EDK9.1SP2 Hard TEMAC version and PLTEMAC version are both 3.0.a PPC clock frequency : 300MHz Kernel : 2.6.21-rc7 , downloaded from grant's git tree some thing near one week ago DMA type: 3 (sg dma) DRE : enabled for TX and RX, (2) CSUM offload is enabled for both of TX and RX tx and rx fifo sizes : 131072 bits the board comes up over NFS root file system completely and without any problems. PC system used for these tests is : CPU P4 Dual Core, 3.4GHz , 2Gigabytes memory, Dual gigabit ethernet port, running linux 2.6.21.3 We have tested the PC system band width and it can easily reach 966mbits/s when connected to the same PC. ( using the same cross cable used for ml403 test) Netperf is compiled with TCP SEND FILE enabled, ( -DHAVE_SENDFILE) (from board to PC) netperf -t TCP_SENDFILE -H 10.10.10.250 -F /boot/zImage.elf -- -m 16384 -s 87380 -S 87380 the measured bandwidth for this test was just 40.66Mbits. It is also true for netperf from PC to board. we do not have any more idea about what we should do to improve the bandwidth. any help or ideas is appreciated... _ Connect to the next generation of MSN Messenger?http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-ussource=wlmailtagline ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ _ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx _ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ ___ Linuxppc-embedded mailing list Linuxppc-embedded@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-embedded
