From: bayi cheng <bayi.cheng@mediatek.com>

The state machine of MTK spi nor controller may be disturbed by some
glitch signals from the relevant BUS during dma read, Although the
possibility of causing the dma read to fail is next to nothing,
However, if error-handling is not implemented, which makes the feature
somewhat risky.

Add an error-handling mechanism here, reset the state machine and
re-read the data when an error occurs.

Signed-off-by: bayi cheng <bayi.cheng@mediatek.com>
---
Change in v2:
  -Add a new function mtk_nor_reset() to reset host state machine.

Change in v1:
  -Reset the state machine when dma read fails and read again.
---
---
 drivers/spi/spi-mtk-nor.c | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/drivers/spi/spi-mtk-nor.c b/drivers/spi/spi-mtk-nor.c
index d167699a1a96..d07f50337f43 100644
--- a/drivers/spi/spi-mtk-nor.c
+++ b/drivers/spi/spi-mtk-nor.c
@@ -80,6 +80,9 @@
 #define MTK_NOR_REG_DMA_FADR		0x71c
 #define MTK_NOR_REG_DMA_DADR		0x720
 #define MTK_NOR_REG_DMA_END_DADR	0x724
+#define MTK_NOR_REG_CG_DIS		0x728
+#define MTK_NOR_SFC_SW_RST		BIT(2)
+
 #define MTK_NOR_REG_DMA_DADR_HB		0x738
 #define MTK_NOR_REG_DMA_END_DADR_HB	0x73c
 
@@ -147,6 +150,15 @@ static inline int mtk_nor_cmd_exec(struct mtk_nor *sp, u32 cmd, ulong clk)
 	return ret;
 }
 
+static void mtk_nor_reset(struct mtk_nor *sp)
+{
+	mtk_nor_rmw(sp, MTK_NOR_REG_CG_DIS, 0, MTK_NOR_SFC_SW_RST);
+	mb(); /* flush previous writes */
+	mtk_nor_rmw(sp, MTK_NOR_REG_CG_DIS, MTK_NOR_SFC_SW_RST, 0);
+	mb(); /* flush previous writes */
+	writel(MTK_NOR_ENABLE_SF_CMD, sp->base + MTK_NOR_REG_WP);
+}
+
 static void mtk_nor_set_addr(struct mtk_nor *sp, const struct spi_mem_op *op)
 {
 	u32 addr = op->addr.val;
@@ -616,7 +628,15 @@ static int mtk_nor_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
 			mtk_nor_set_addr(sp, op);
 			return mtk_nor_read_pio(sp, op);
 		} else {
-			return mtk_nor_read_dma(sp, op);
+			ret = mtk_nor_read_dma(sp, op);
+			if (unlikely(ret)) {
+				/* Handle rare bus glitch */
+				mtk_nor_reset(sp);
+				mtk_nor_setup_bus(sp, op);
+				return mtk_nor_read_dma(sp, op);
+			}
+
+			return ret;
 		}
 	}
 
-- 
2.18.0

On Wed, 07 Dec 2022 13:54:35 +0800, Bayi Cheng wrote:
> The state machine of MTK spi nor controller may be disturbed by some
> glitch signals from the relevant BUS during dma read, Although the
> possibility of causing the dma read to fail is next to nothing,
> However, if error-handling is not implemented, which makes the feature
> somewhat risky.
> 
> Add an error-handling mechanism here, reset the state machine and
> re-read the data when an error occurs.
> 
> [...]

Applied to

   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git for-next

Thanks!

[1/1] spi: spi-mtk-nor: Add recovery mechanism for dma read timeout
      commit: 8330e9e8269bb76dd502e84efb5f351016512cf8

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark

Il 07/12/22 06:54, Bayi Cheng ha scritto:
> From: bayi cheng <bayi.cheng@mediatek.com>
> 
> The state machine of MTK spi nor controller may be disturbed by some
> glitch signals from the relevant BUS during dma read, Although the
> possibility of causing the dma read to fail is next to nothing,
> However, if error-handling is not implemented, which makes the feature
> somewhat risky.
> 
> Add an error-handling mechanism here, reset the state machine and
> re-read the data when an error occurs.
> 
> Signed-off-by: bayi cheng <bayi.cheng@mediatek.com>

Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>