89712 is based on 7312 ; there was a bug in 7312 first masks, so I suspect there has been the same problem on the cs89712 the 7312 was usable only with rev-C version of the chip can you verify the revision of your chip (no clear relation with the board revision...) according to the marking on the chip and information form cirrus and make sure you have a working processor giving you access to SDRAM with MMU off
Xavier S�bastien Judenherc a �crit : > > Hello, > after many tests and off-list discussions, it appears that the problem > is in on-chip sram. > Using hermit, it is possible do download data to dram while a download > to dlbuf (addr="buf" when using the download command) generates a crc > mismatch error - dlbuf is located in onchip-sram when booting with JP31 > open. > I did the following test: catch the "download buf 8192 0x2c963992" > command from hermit -v ..., then in hermit with terminal otpion: > "download 0xc0000000 8192 0x2c963992" and from an other console, I send > the first 8192 bytes to ttyS0 -> no error. When I do exactly the same > but "buf" in stead of 0xc0000000 I get a CRC mismatch. > > I have tested different SDRAM refresh rate ... no effect. > > Reducing the cpu speed to 37MHz in target/cs89712/boot.S does not change > anything as well: > /* mov r1, #0x6 */ /* CLKCTL_73 in SYSCON3 */ > mov r1, #0x2 /* 37 MHz */ > add r2, r0, #0x2200 /* address of SYSCON3 in r2 */ > str r1, [r2] > while I can check that register have the correct value in hermit > > hermit:frob> p 0x80002200 > *0x80002200 == 0x00000002 > > does it mean that the spu speed is actually 37MHz ? > > Am I the only arm-linux user with a rev A board on the list ? > > desperately, s�bastien > > _______________________________________________ > http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm > Please visit the above address for information on this list. _______________________________________________ http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm Please visit the above address for information on this list.
