Hi Saeed, There is no any chance to read the PCIe register from user space. Because when it happens, system is just like hang without any response. The only chance is to modify it in the drivers. If you want to see some special register, you can tell me, i would modify the driver to print it out. Below is the last register information from devmem2. Sorry for we don't have other PCIe device to test. And this issue would happens when using the Marvell BSP 4.2.2. The result is all the same as sata_mv.c (Dead in deactivateEdma, keep looping with the MV_SATA_UNRECOVERABLE_COMMUNICATION_ERROR event) Thanks. ------------------ Value at address 0x22000 (0x40000000): 0x28126 Value at address 0x24000 (0x40000000): 0x41ED Value at address 0x26000 (0x40000000): 0xE1A00000 Value at address 0x28000 (0x40000000): 0xE1A01008 ------------------------ Value at address 0x22030 (0x40000030): 0x280D7 Value at address 0x24030 (0x40000030): 0x50803 Value at address 0x26030 (0x40000030): 0xEB09FE5F Value at address 0x28030 (0x40000030): 0x12FFF10 ... cannot deactivate Edma .. keep looping. Best Regards, KueiHuan-Chen 2009/7/22 saeed bishara <saeed.bishara@xxxxxxxxx>: > Hi, > looks to me something wrong happened in PCIe link level, and any > register read ends up with 0xFFFFFFFF. can you please use devmem2 tool > and read any register (e.g. 0x22000) of the device? > > also, can you test any other PCIe device? > > does this issue happens when using the Marvell BSP 4.2.2? > saeed > > On Wed, Jul 22, 2009 at 10:04 AM, KueiHuan Chen<kueihuan.chen@xxxxxxxxx> wrote: >> Dear Mark & Saeed, >> >> 2009/7/22 Mark Lord <mlord@xxxxxxxxx>: >>> >>> Are you using a discrete 7042 chip on a PCIe bus >>> along with a 6281 chip ?? >>> Or are you simply using the built-in SATA controller inside the 6281 ? >>> >> >> We use a standalone 7042 chip on a PCIe bus with 6281 platform. And >> we choose four channels from 7042 to build a software raid 5. We had >> experimented that even though we disable the sata channels of 6281 >> soc, the problem is still remain. >> >>> In either event, Marvell did not deem to share *any* technical info >>> about the 6281 with me, so I really cannot help you. >>> >>> Saeed Bishara from Marvell did everything to do with that chip. >>> >>> >>> .. >>> >>> I think the messages that happened before those above are the important >>> ones, >>> and you did not show those messages here. >> >> There is no any messages before the "unable to disable eDMA". It >> just freeze for a while and then output these messages. I had tried to >> use the sata driver from Marvell BSP driver 4.2.2. The problem is the >> same. The BSP are also cannot disable eDMA. Then we unmount this chip >> from that board and remount it into the other platform(powerpc). It >> works fine without any problem ... So we suspicion the problem may due >> to 7042 work on 6281 platform. >> >> If you have any ideas , please kindly tell us. We would try it ! >> >> Thanks! >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-ide" in >> the body of a message to majordomo@xxxxxxxxxxxxxxx >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > -- To unsubscribe from this list: send the line "unsubscribe linux-ide" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html