cfi_cmdset_0001.c: Excessive erase suspends

Joakim Tjernlund Joakim.Tjernlund at transmode.se
Sat Apr 19 13:01:20 EDT 2008



> -----Original Message-----
> From: Anders Grafström [mailto:grfstrm at users.sourceforge.net]
> Sent: den 19 april 2008 15:47
> To: Joakim Tjernlund; 'Jared Hulbert'; 'Alexey Korolev'; 'Linux-MTD Mailing List'
> Subject: Re: cfi_cmdset_0001.c: Excessive erase suspends
> 
> Joakim Tjernlund wrote:
> > Don't you need to check that the chip is suspended before you do any
> > new operation?
> 
> You're right.
> 
> Here's the improved version;
> 
> static void erase_suspend(unsigned int loops)
> {
> 	volatile unsigned int *flash = (unsigned int *)0x41040000;
> 	unsigned int i;
> 	unsigned int suspends = 0;
> 	unsigned int words = 2;
> 	unsigned int timeout;
> 
> 	/* Make sure there's something to erase */
> 	for (i = 0; i < 0x10000; i++) {
> 		if (flash[i] == 0xffffffff) {
> 			flash[i] = 0x00400040;
> 			flash[i] = 0;
> 			while ((flash[i] & 0x00800080) != 0x00800080);
> 		}
> 	}
> 
> 	/* Clear status */
> 	flash[0] = 0x00500050;
> 
> 	/* Erase */
> 	flash[0] = 0x00200020;
> 	flash[0] = 0x00d000d0;
> 
> 	while (1) {
> 		/* Suspend */
> 		flash[0] = 0x00b000b0;
> 		flash[0] = 0x00700070;
> 		timeout = 100000;
> 		while ((flash[0] & 0x00800080) != 0x00800080 && --timeout);
> 		if (!timeout)
> 			printf("suspend timeout\n");
> 		suspends++;
> 
> 		/* Short delay */
> 		for (i = 0; i < 1000; i++);
> 
> 		/* Resume */
> 		flash[0] = 0x00d000d0;
> 		flash[0] = 0x00700070;
> 
> 		/* Erase done ? */
> 		if ((flash[0] & 0x00800080) == 0x00800080) {
> 			printf("\nstatus %08x\n", flash[0]);
> 			break;
> 		}
> 
> 		/* Short delay */
> 		for (i = 0; i < loops; i++);
> 	}
> 
> 	/* Read array */
> 	flash[0] = 0x00ff00ff;
> 
> 	/* Show the first 2 failed words */
> 	for (i = 0; i < 0x10000; i++) {
> 		if (flash[i] != 0xffffffff) {
> 			printf("%08x\n", flash[i]);
> 			if (--words == 0)
> 				break;
> 		}
> 	}
> 
> 	printf("%d suspends\n", suspends);
> }
> 
> void test_flash_erase_suspend(void)
> {
> 	unsigned int loops;
> 
> 	for (loops = 6000; loops > 2000; loops -= 500) {
> 		printf("\n%d loops", loops);
> 		erase_suspend(loops);
> 	}
> }
> 
> 
> It produced this output at one occasion:
> 
> 6000 loops
> status 00800080
> ffff0000
> ffff0000
> 4783 suspends
> 
> 5500 loops
> status 00800080
> 5508 suspends
> 
> 5000 loops
> status 008000a0
> 6394 suspends
> 
> 4500 loops
> status 00a000a0
> fffffeff
> fffffffe
> 5169 suspends
> 
> 4000 loops
> status 00a000a0
> fffffeff
> fffffeff
> 5378 suspends
> 
> 3500 loops
> status 00a000a0
> aaaaaaaa
> aaaaaaaa
> 5715 suspends
> 
> 3000 loops
> status 00a000a0
> 55555555
> 55555555
> 6142 suspends
> 
> 2500 loops
> status 00a000a0
> 00000000
> 00000000
> 6929 suspends
> 
> It locks up when I drop it down to 2000 or less iterations in the loop between resume and suspend.
> I think I need to do more testing with properly timed delays.
> 
> Is there anything else that I have missed?

Perhaps, you should make "i" a volatile. That said I know that Intel
had erase suspend bugs in their early strata flash, that’s why there is
a define in cmdset_0001 that disables them.

Maybe you got buggy chips?
oh, just remembered, strata chips needs to have a stable address bus before
CS is pulled. We had add a small delay on CS in the memory controller, otherwise
they would fail, especially when cold. 

> 
> /Anders
> 





More information about the linux-mtd mailing list