hi exports:
recently,we had researched the dm81XX series,included dm8168,dm8148.We also have the ti evm board of dm8168 and dm8148. In the dm8148 EVM board debugging, we mount the rootfs by nfs,and the kernel started ,the nfs rootfs mounted successfully. When we executed the cmem mould the to load the code file (c674 and ducati m3 which compiled by the CCS4.2 and put in the NFS share rootfs), but the kernel read function holded(just read few bytes successfully) ,and then the uart show the log : nfs: server 192.168.18.18 not responding, still trying nfs: server 192.168.18.18 not responding, still trying
we used many ways to prove the cmem mould and the nfs sever are ok, such as:
(1) We used the ethernet cable to link the LINUX host and the EVM board directly, the uImage loaded ok(by tftp server ),and the rootfs mounted ok (by nfs ), then we load the code file,the phenomena is like before.
(2)we used the mkfs tool to make the rootfs image,and wrote the rootfs image into the nand flash,then modifyed the uboot args like that
setenv bootargs 'console=ttyO0,115200n8 noinitrd root=/dev/mtdblock4 rw rootfstype=jffs2 mem=160M earlyprintk rootdelay=4'
then,we boot the sys and load the code file,erveryting is ok.
(3)We used the dm8168 EVM to test the nfs sever, the same rootfs and related IP addr, and the same mode of loadding code file ,but everyting is OK
(4)We wrote a simple code to realize writing&readding files function in the rootfs which is mount by nfs share,and the volume of file is bigger than code file,but everything is ok.
(5)last year ,i had used the centaurus(dm8148) evm board (not have the LCD board)which is made by MISTRAL,but that board just had one RJ45 Ethernet. But now,our new dm8148 evm board have two RJ45 Ethernet. When we used the prebulid uboot and uImage in ezsdk,and booted in related mode,just one RJ45 Ethernet is available(J14).We think it might the reason of EMAC Switch and CPSW Setting,but we knew little about these. We just suspected the NFS err is related with these.