Re: [Ocf-linux-users] infinite wait in mv_cesa (Marvell Kirkwood) icw cryptotest
Brought to you by:
david-m
Menu
▾
▴
Re: [Ocf-linux-users] infinite wait in mv_cesa (Marvell Kirkwood) icw cryptotest
|
From: Ronen S. <rsh...@ma...> - 2010-08-29 15:57:53
|
The idea was that accessing a reg while using a base as a define is faster then having the base as a variable... -----Original Message----- From: Harro Haan [mailto:hr...@gm...] Sent: Tuesday, August 24, 2010 12:30 PM To: dav...@mc...; Ronen Shitrit Cc: Pat...@bd...; ocf...@li... Subject: Re: infinite wait in mv_cesa (Marvell Kirkwood) icw cryptotest Problem solved! INTER_REGS_BASE (in mvSysHwConfig.h) had the value 0xFEE00000. This is the same as KIRKWOOD_PCIE_IO_VIRT_BASE in linux-2.6.33/arch/arm/mach-kirkwood/include/mach/kirkwood.h. For my configuration it should be KIRKWOOD_REGS_VIRT_BASE, which is 0xfed00000. Why does the define INTER_REGS_BASE exist? Would it not be cleaner to use “cesa_device.reg” instead (set by platform_get_resource_byname() in cesa_ocf_drv.c)? Regards, Harro On 23 August 2010 16:47, Harro Haan <hr...@gm...> wrote: > Some additional info, > > When I define CESA_OCF_POLLING in cesa_ocf_drv.c everything works fine. > > When I do "cat /proc/interrupts" (with #undef CESA_OCF_POLLING), I can > see that no cesa interrupts occurred. When I look at the debug output > (see previous mail) I do not see the line "cesa_interrupt_handler()" > either. > > So for my configuration something is wrong/missing with my interrupts. > Any hints? > > Thanks, > > Harro > > On 20 August 2010 16:52, Harro Haan <hr...@gm...> wrote: >> Hello David, >> >> I’m trying to get this driver working with a 3.6.33 kernel on my >> OpenRD-Ultimate 88F6281 development board, but the cryptotest program >> is waiting forever. >> >> Debug output for mv_cesa: >> ------------------------------------ >> root@DB88FXX81:/ ./cryptotest -a aes >> cryptodev_open() >> cryptodev_ioctl(cmd=c0046364 arg=beea0954) >> cryptodev_ioctl(CRIOGET) >> cryptodev_ioctl(cmd=c030636a arg=beea09b4) >> cryptodev_ioctl(CIOCGSESSION2) >> cryptodev_ioctl(CIOCGSESSION2) - no mac >> crypto/ocf/crypto.c,389: DRIVER_LOCK() >> crypto/ocf/crypto.c,413: DRIVER_UNLOCK() >> cesa_ocf_newsession() >> crypto/ocf/kirkwood/cesa_ocf_drv.c,875: new session 1 >> crypto/ocf/kirkwood/cesa_ocf_drv.c,896: (1) AES CBC >> crypto/ocf/kirkwood/cesa_ocf_drv.c,906: key length 16 >> crypto/ocf/crypto.c,415: DRIVER_LOCK() >> crypto/ocf/crypto.c,425: DRIVER_UNLOCK() >> csecreate() >> cseadd() >> cryptodev_ioctl(cmd=c01c6367 arg=beea0998) >> cryptodev_ioctl(CIOCCRYPT) >> csefind() >> cryptodev_op() >> crypto_dispatch() >> crypto/ocf/crypto.c,815: Q_LOCK() >> crypto/ocf/crypto.c,839: Q_UNLOCK() >> crypto_invoke() >> cesa_ocf_process() >> crypto/ocf/kirkwood/cesa_ocf_drv.c,425: handle UIO. >> crypto/ocf/kirkwood/cesa_ocf_drv.c,440: buf 0-> addr ded17500, size 10 >> crypto/ocf/kirkwood/cesa_ocf_drv.c,480: cipher encrypt >> crypto/ocf/kirkwood/cesa_ocf_drv.c,491: IV from USER (offset 0) >> crypto/ocf/kirkwood/cesa_ocf_drv.c,508: don't copy the IV back to the buffer >> >> crypto/ocf/kirkwood/cesa_ocf_drv.c,596: Sending Action: >> crypto/ocf/kirkwood/cesa_ocf_drv.c,597: IV from user: 1. IV offset 0 >> crypto/ocf/kirkwood/cesa_ocf_drv.c,598: crypt offset 10 len 10 >> crypto/ocf/kirkwood/cesa_ocf_drv.c,599: Auth offset 0 len 0 >> crypto/ocf/kirkwood/cesa_ocf_drv.c,600: set digest in offset 0 . >> SRC BUFFER: pMbuf=dfa4bc24, numFrags=2, mbufSize=32 >> # 0. bufVirt=df8ee020, bufSize=16 >> df8ee020: 33 35 38 36 74 34 69 6f 68 33 75 21 75 36 34 39 >> # 1. bufVirt=ded17500, bufSize=16 >> ded17500: 6e 33 38 39 34 36 6f 74 38 69 74 39 31 65 75 6e >> crypto/ocf/crypto.c,841: Q_LOCK() >> crypto/ocf/crypto.c,867: Q_UNLOCK() >> cryptodev_op about to WAIT >> >> ------------------------------------ >> >> Debug output for cryptosoft (which is working just fine): >> ------------------------------------ >> root@DB88FXX81:/ ./cryptotest -a aes >> cryptodev_open() >> cryptodev_ioctl(cmd=c0046364 arg=bea58954) >> cryptodev_ioctl(CRIOGET) >> cryptodev_ioctl(cmd=c030636a arg=bea589b4) >> cryptodev_ioctl(CIOCGSESSION2) >> cryptodev_ioctl(CIOCGSESSION2) - no mac >> crypto/ocf/crypto.c,389: DRIVER_LOCK() >> crypto/ocf/crypto.c,413: DRIVER_UNLOCK() >> swcr_newsession() >> swcr_newsession crypto_alloc_*blkcipher(cbc(aes), 0x0) >> swcr_newsession cbc(aes) cipher is async >> swcr_newsession key:cri->cri_klen=128,(cri->cri_klen + 7)/8=16 >> 0x32 0x61 0x38 0x6f 0x21 0x32 0x61 0x61 >> 0x21 0x35 0x75 0x21 0x69 0x30 0x65 0x21 >> crypto/ocf/crypto.c,415: DRIVER_LOCK() >> crypto/ocf/crypto.c,425: DRIVER_UNLOCK() >> csecreate() >> cseadd() >> cryptodev_ioctl(cmd=c01c6367 arg=bea58998) >> cryptodev_ioctl(CIOCCRYPT) >> csefind() >> cryptodev_op() >> crypto_dispatch() >> crypto/ocf/crypto.c,815: Q_LOCK() >> crypto/ocf/crypto.c,839: Q_UNLOCK() >> crypto_invoke() >> swcr_process() >> swcr_process_req() >> crypto OP success 0 >> crypto_done() >> crypto/ocf/crypto.c,1156: Q_LOCK() >> crypto/ocf/crypto.c,1158: Q_UNLOCK() >> cryptodev_cb() >> crypto/ocf/crypto.c,841: Q_LOCK() >> crypto/ocf/crypto.c,867: Q_UNLOCK() >> cryptodev_op about to WAIT >> cryptodev_op finished WAITING error=0 >> cryptodev_ioctl(cmd=c01c6367 arg=bea58998) >> cryptodev_ioctl(CIOCCRYPT) >> csefind() >> cryptodev_op() >> crypto_dispatch() >> crypto/ocf/crypto.c,815: Q_LOCK() >> crypto/ocf/crypto.c,839: Q_UNLOCK() >> crypto_invoke() >> swcr_process() >> swcr_process_req() >> crypto OP success 0 >> crypto_done() >> crypto/ocf/crypto.c,1156: Q_LOCK() >> crypto/ocf/crypto.c,1158: Q_UNLOCK() >> cryptodev_cb() >> crypto/ocf/crypto.c,841: Q_LOCK() >> crypto/ocf/crypto.c,867: Q_UNLOCK() >> cryptodev_op about to WAIT >> cryptodev_op finished WAITING error=0 >> cryptodev_ioctl(cmd=40046366 arg=bea589cc) >> cryptodev_ioctl(CIOCFSESSION) >> csefind() >> csedelete() >> csefree() >> crypto_freesession() >> crypto/ocf/crypto.c,450: DRIVER_LOCK() >> crypto/ocf/crypto.c,468: DRIVER_UNLOCK() >> swcr_freesession() >> crypto/ocf/crypto.c,471: DRIVER_LOCK() >> crypto/ocf/crypto.c,481: DRIVER_UNLOCK() >> 0.089 sec, cryptodev_release() >> 2 aes crypts, 16 bytes, 358 byte/sec, 0.0 Mb/sec >> root@DB88FXX81:/ >> ------------------------------------ >> >> I am following this thread as well: >> “[Ocf-linux-users] OCF driver for Marvell Kirkwood (88F6xxx) Processors” >> http://sourceforge.net/mailarchive/forum.php?thread_name=8D145DCD0F55B543949BE94258003A200DF9AD4EA6%40exchange12.bdt-rw.de&forum_name=ocf-linux-users >> >> Attached are my patches (fixes as described in the thread above) and >> series file. >> >> Do you have an idea why the system is hanging/waiting forever? >> >> Thanks, >> >> Harro >> > |
