Hi Stefan,
Le Mon, 27 Jul 2015 18:42:41 +0200, Stefan Agner <ste...@agner.ch> a
écrit :
> This driver supports Freescale NFC (NAND flash controller) found on
> Vybrid (VF610), MPC5125, MCF54418 and Kinetis K70. The driver has
> been tested on 8-bit and 16-bit NAND interface and supports ONFI
> parameter page reading.
>
> [...]
>
> diff --git a/drivers/mtd/nand/vf610_nfc.c b/drivers/mtd/nand/vf610_nfc.c
> new file mode 100644
> index 0000000..0da500e
> --- /dev/null
> +++ b/drivers/mtd/nand/vf610_nfc.c
> @@ -0,0 +1,640 @@
> [...]
... about line 708:
> + err = devm_request_irq(nfc->dev, irq, vf610_nfc_irq, 0, DRV_NAME, mtd);
> + if (err) {
> + dev_err(nfc->dev, "Error requesting IRQ!\n");
> + goto error;
> + }
> +
> + vf610_nfc_init_controller(nfc);
The call above is too early: vf610_nfc_init_controller() will test
for (nfc->chip.options & NAND_BUSWIDTH_16) but this option bit is only
set once nand_scan_ident() below has been run.
This has the effect that even when the DT node specifies a 16-bit wide
bus, the controller is configured for 8-bit mode at this point, which of
course causes read failures. I've experienced this with a Vybrid SoC
and a Micron MT29F4G16ABADAH4 16-bit NAND.
> + /* first scan to find the device and get the page size */
> + if (nand_scan_ident(mtd, 1, NULL)) {
> + err = -ENXIO;
> + goto error;
> + }
Placing the call to vf610_nfc_init_controller() here, after the call
to nand_scan_ident() rather than before it, fixed the issue for me.
Cordialement,
Albert ARIBAUD
3ADEV
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
