Re: Huge performance degradation for UDP between 2.4.17 and 2.6
Le 05/08/2012 10:28, Eric Dumazet a écrit : On Sun, 2012-08-05 at 10:16 +0200, LEROY christophe wrote: Le 02/08/2012 16:13, Eric Dumazet a écrit : On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. Thanks for your answer. Yes I have netfilter as I need it. However, I tried without it and still need about 37 seconds to send the 16 packets I was sending in 11 seconds with 2.4.17 I don't think there is any problem with size of the kernel. I still have plenty of memory available. I believe you misunderstood me. I was referring to cpu caches ( dcache & icache ) All debugging is turned off, and I'm not using SLUB but SLOB. I have 32Mbytes of RAM. Would SLUB be more performant than SLOB ? I never used SLOB I cannot comment Please provide (on 3.4.7) cat /proc/cpuinfo lsmod dmesg Ok, I have recompiled with SLUB. Find attached cpuinfo, lsmod and dmesg. I do not have any modules loaded. Module Size Used byNot tainted processor : 0 cpu : 8xx clock : 132.00MHz revision: 0.0 (pvr 0050 ) bogomips: 16.50 timebase: 825 platform: CMPC885 model : MIAE Memory : 128 MB [0.00] Using CMPC885 machine description [0.00] Linux version 3.4.7-s3k-3.8.3+-svn2796 (root@localhost.localdomain) (gcc version 4.4.4 (GCC) ) #19 PREEMPT Sun Aug 5 06:43:31 CEST 2012 [0.00] Top of RAM: 0x800, Total RAM: 0x800 [0.00] Memory hole size: 0MB [0.00] Zone PFN ranges: [0.00] DMA 0x -> 0x8000 [0.00] Normal empty [0.00] Movable zone start PFN for each node [0.00] Early memory PFN ranges [0.00] 0: 0x -> 0x8000 [0.00] On node 0 totalpages: 32768 [0.00] free_area_init_node: node 0, pgdat c03f4be8, node_mem_map c0433000 [0.00] DMA zone: 256 pages used for memmap [0.00] DMA zone: 0 pages reserved [0.00] DMA zone: 32512 pages, LIFO batch:7 [0.00] MMU: Allocated 72 bytes of context maps for 16 contexts [0.00] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768 [0.00] pcpu-alloc: [0] 0 [0.00] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 32512 [0.00] Kernel command line: console=ttyCPM0,115200N8 ip=172.25.231.39:172.25.231.59::255.0.0.0:miae:eth0:off [0.00] PID hash table entries: 512 (order: -1, 2048 bytes) [0.00] Dentry cache hash table entries: 16384 (order: 4, 65536 bytes) [0.00] Inode-cache hash table entries: 8192 (order: 3, 32768 bytes) [0.00] Memory: 125432k/131072k available (3972k kernel code, 5640k reserved, 140k data, 178k bss, 488k init) [0.00] Kernel virtual memory layout: [0.00] * 0xfffdf000..0xf000 : fixmap [0.00] * 0xfde0..0xfe00 : consistent mem [0.00] * 0xfddf6000..0xfde0 : early ioremap [0.00] * 0xc900..0xfddf6000 : vmalloc & ioremap [0.00] SLUB: Genslabs=14, HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=1 [0.00] NR_IRQS:512 nr_irqs:512 16 [0.00] Decrementer Frequency = 0x7de290 [0.00] time_init: decrementer frequency = 8.25 MHz [0.00] time_init: processor frequency = 132.00 MHz [0.00] clocksource: timebase mult[79364d93] shift[24] registered [0.00] clockevent: decrementer mult[21cac08] shift[32] cpu[0] [0.00] console [ttyCPM0] enabled [0.144102] pid_max: default: 32768 minimum: 301 [0.149489] Mount-cache hash table entries: 512 [0.175086] devtmpfs: initialized [0.178575] device: 'platform': device_add [0.179079] bus: 'platform': registered [0.179451] bus: 'cpu': registered [0.179551] device: 'cpu': device_add [0.181718] NET: Registered protocol family 16 [0.186374] device class 'bdi': registering [0.188054] device class 'gpio': registering [0.189423] device class 'tty': registering [0.191855] bus: 'spi': registered [0.191932] device class 'spi_master': registering [0.193842] gpiochip_find_base: found new base at 1008 [0.193974] device: 'gpiochip1008': device_add [0.195387] gpiochip_add: registered GPIOs 1008 to 1023 on device: /soc@ff00/cpm@9c0/gpio-controller@950 [0.205830] gpiochip_find_base: found new base at 976 [
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
Le 05/08/2012 10:28, Eric Dumazet a écrit : On Sun, 2012-08-05 at 10:16 +0200, LEROY christophe wrote: Le 02/08/2012 16:13, Eric Dumazet a écrit : On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. Thanks for your answer. Yes I have netfilter as I need it. However, I tried without it and still need about 37 seconds to send the 16 packets I was sending in 11 seconds with 2.4.17 I don't think there is any problem with size of the kernel. I still have plenty of memory available. I believe you misunderstood me. I was referring to cpu caches ( dcache icache ) All debugging is turned off, and I'm not using SLUB but SLOB. I have 32Mbytes of RAM. Would SLUB be more performant than SLOB ? I never used SLOB I cannot comment Please provide (on 3.4.7) cat /proc/cpuinfo lsmod dmesg Ok, I have recompiled with SLUB. Find attached cpuinfo, lsmod and dmesg. I do not have any modules loaded. Module Size Used byNot tainted processor : 0 cpu : 8xx clock : 132.00MHz revision: 0.0 (pvr 0050 ) bogomips: 16.50 timebase: 825 platform: CMPC885 model : MIAE Memory : 128 MB [0.00] Using CMPC885 machine description [0.00] Linux version 3.4.7-s3k-3.8.3+-svn2796 (root@localhost.localdomain) (gcc version 4.4.4 (GCC) ) #19 PREEMPT Sun Aug 5 06:43:31 CEST 2012 [0.00] Top of RAM: 0x800, Total RAM: 0x800 [0.00] Memory hole size: 0MB [0.00] Zone PFN ranges: [0.00] DMA 0x - 0x8000 [0.00] Normal empty [0.00] Movable zone start PFN for each node [0.00] Early memory PFN ranges [0.00] 0: 0x - 0x8000 [0.00] On node 0 totalpages: 32768 [0.00] free_area_init_node: node 0, pgdat c03f4be8, node_mem_map c0433000 [0.00] DMA zone: 256 pages used for memmap [0.00] DMA zone: 0 pages reserved [0.00] DMA zone: 32512 pages, LIFO batch:7 [0.00] MMU: Allocated 72 bytes of context maps for 16 contexts [0.00] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768 [0.00] pcpu-alloc: [0] 0 [0.00] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 32512 [0.00] Kernel command line: console=ttyCPM0,115200N8 ip=172.25.231.39:172.25.231.59::255.0.0.0:miae:eth0:off [0.00] PID hash table entries: 512 (order: -1, 2048 bytes) [0.00] Dentry cache hash table entries: 16384 (order: 4, 65536 bytes) [0.00] Inode-cache hash table entries: 8192 (order: 3, 32768 bytes) [0.00] Memory: 125432k/131072k available (3972k kernel code, 5640k reserved, 140k data, 178k bss, 488k init) [0.00] Kernel virtual memory layout: [0.00] * 0xfffdf000..0xf000 : fixmap [0.00] * 0xfde0..0xfe00 : consistent mem [0.00] * 0xfddf6000..0xfde0 : early ioremap [0.00] * 0xc900..0xfddf6000 : vmalloc ioremap [0.00] SLUB: Genslabs=14, HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=1 [0.00] NR_IRQS:512 nr_irqs:512 16 [0.00] Decrementer Frequency = 0x7de290 [0.00] time_init: decrementer frequency = 8.25 MHz [0.00] time_init: processor frequency = 132.00 MHz [0.00] clocksource: timebase mult[79364d93] shift[24] registered [0.00] clockevent: decrementer mult[21cac08] shift[32] cpu[0] [0.00] console [ttyCPM0] enabled [0.144102] pid_max: default: 32768 minimum: 301 [0.149489] Mount-cache hash table entries: 512 [0.175086] devtmpfs: initialized [0.178575] device: 'platform': device_add [0.179079] bus: 'platform': registered [0.179451] bus: 'cpu': registered [0.179551] device: 'cpu': device_add [0.181718] NET: Registered protocol family 16 [0.186374] device class 'bdi': registering [0.188054] device class 'gpio': registering [0.189423] device class 'tty': registering [0.191855] bus: 'spi': registered [0.191932] device class 'spi_master': registering [0.193842] gpiochip_find_base: found new base at 1008 [0.193974] device: 'gpiochip1008': device_add [0.195387] gpiochip_add: registered GPIOs 1008 to 1023 on device: /soc@ff00/cpm@9c0/gpio-controller@950 [0.205830] gpiochip_find_base: found new base at 976 [
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
On Sun, 2012-08-05 at 10:16 +0200, LEROY christophe wrote: > Le 02/08/2012 16:13, Eric Dumazet a écrit : > > On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: > >> Hi > >> > >> I'm having a big issue with UDP. Using a powerpc board (MPC860). > >> > >> With our board running kernel 2.4.17, I'm able to send 16 voice > >> packets (UDP, 96 bytes per packet) in 11 seconds. > >> With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I > >> need 55 seconds to send the same amount of packets. > >> > >> > >> Is there anything to tune in order to get same output rate as with > >> Kernel 2.4 ? > > kernel size is probably too big for your old / slow cpu. > > > > Maybe you added too many features on your 3.4.7 kernel. (netfilter ? > > SLUB debugging ...) > > > > Its hard to say, 2.4.17 had less features and was faster. > > > > Thanks for your answer. > Yes I have netfilter as I need it. However, I tried without it and still > need about 37 seconds to send the 16 packets I was sending in 11 > seconds with 2.4.17 > > I don't think there is any problem with size of the kernel. I still have > plenty of memory available. > I believe you misunderstood me. I was referring to cpu caches ( dcache & icache ) > All debugging is turned off, and I'm not using SLUB but SLOB. > I have 32Mbytes of RAM. Would SLUB be more performant than SLOB ? I never used SLOB I cannot comment Please provide (on 3.4.7) cat /proc/cpuinfo lsmod dmesg -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
Le 02/08/2012 16:13, Eric Dumazet a écrit : On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. Thanks for your answer. Yes I have netfilter as I need it. However, I tried without it and still need about 37 seconds to send the 16 packets I was sending in 11 seconds with 2.4.17 I don't think there is any problem with size of the kernel. I still have plenty of memory available. All debugging is turned off, and I'm not using SLUB but SLOB. I have 32Mbytes of RAM. Would SLUB be more performant than SLOB ? Christophe -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
Le 02/08/2012 16:13, Eric Dumazet a écrit : On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. Thanks for your answer. Yes I have netfilter as I need it. However, I tried without it and still need about 37 seconds to send the 16 packets I was sending in 11 seconds with 2.4.17 I don't think there is any problem with size of the kernel. I still have plenty of memory available. All debugging is turned off, and I'm not using SLUB but SLOB. I have 32Mbytes of RAM. Would SLUB be more performant than SLOB ? Christophe -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
On Sun, 2012-08-05 at 10:16 +0200, LEROY christophe wrote: Le 02/08/2012 16:13, Eric Dumazet a écrit : On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. Thanks for your answer. Yes I have netfilter as I need it. However, I tried without it and still need about 37 seconds to send the 16 packets I was sending in 11 seconds with 2.4.17 I don't think there is any problem with size of the kernel. I still have plenty of memory available. I believe you misunderstood me. I was referring to cpu caches ( dcache icache ) All debugging is turned off, and I'm not using SLUB but SLOB. I have 32Mbytes of RAM. Would SLUB be more performant than SLOB ? I never used SLOB I cannot comment Please provide (on 3.4.7) cat /proc/cpuinfo lsmod dmesg -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
On 08/02/2012 06:27 AM, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? Have you tried profiling it to see where the time is being spent? Also turning off any unnecessary features, debugging, etc? Chris -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: > Hi > > I'm having a big issue with UDP. Using a powerpc board (MPC860). > > With our board running kernel 2.4.17, I'm able to send 16 voice > packets (UDP, 96 bytes per packet) in 11 seconds. > With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I > need 55 seconds to send the same amount of packets. > > > Is there anything to tune in order to get same output rate as with > Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
On Thu, 2012-08-02 at 14:27 +0200, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? kernel size is probably too big for your old / slow cpu. Maybe you added too many features on your 3.4.7 kernel. (netfilter ? SLUB debugging ...) Its hard to say, 2.4.17 had less features and was faster. -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: Huge performance degradation for UDP between 2.4.17 and 2.6
On 08/02/2012 06:27 AM, leroy christophe wrote: Hi I'm having a big issue with UDP. Using a powerpc board (MPC860). With our board running kernel 2.4.17, I'm able to send 16 voice packets (UDP, 96 bytes per packet) in 11 seconds. With the same board running either Kernel 2.6.35.14 or Kernel 3.4.7, I need 55 seconds to send the same amount of packets. Is there anything to tune in order to get same output rate as with Kernel 2.4 ? Have you tried profiling it to see where the time is being spent? Also turning off any unnecessary features, debugging, etc? Chris -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/