I have encountered an interrupt storm during the eMMC chip probing (and
the chip finally didn't get detected). It turned out that U-Boot left
the SDHI DMA interrupts enabled while the Linux driver didn't use those.
Masking those interrupts in renesas_sdhi_internal_dmac_request_dma() gets
rid of both issues...
Signed-off-by: Sergei Shtylyov <sergei.shtylyov@xxxxxxxxxxxxxxxxxx>
---
The patch is against Ulf Hansson's 'mmc.git' repo's 'fixes' branch.
Changes in version 2:
- fixed #define INFO{1|2}_RESERVED_BITS, rename them to INFO{1|2}_MASK_CLEAR,
and moved them higher;
- shortened the comment before masking the DMAC interrupts;
- fixed up the excess whitespace in the patch dsescription.
drivers/mmc/host/renesas_sdhi_internal_dmac.c | 8 ++++++++
1 file changed, 8 insertions(+)
Index: renesas/drivers/mmc/host/renesas_sdhi_internal_dmac.c
===================================================================
--- renesas.orig/drivers/mmc/host/renesas_sdhi_internal_dmac.c
+++ renesas/drivers/mmc/host/renesas_sdhi_internal_dmac.c
@@ -49,10 +49,12 @@
/* DM_CM_INFO1 and DM_CM_INFO1_MASK */
#define INFO1_CLEAR 0
+#define INFO1_MASK_CLEAR GENMASK_ULL(31, 0)
#define INFO1_DTRANEND1 BIT(17)
#define INFO1_DTRANEND0 BIT(16)
/* DM_CM_INFO2 and DM_CM_INFO2_MASK */
+#define INFO2_MASK_CLEAR GENMASK_ULL(31, 0)
#define INFO2_DTRANERR1 BIT(17)
#define INFO2_DTRANERR0 BIT(16)
@@ -236,6 +238,12 @@ renesas_sdhi_internal_dmac_request_dma(s
{
struct renesas_sdhi *priv = host_to_priv(host);
+ /* Disable DMAC interrupts, we don't use them */
+ renesas_sdhi_internal_dmac_dm_write(host, DM_CM_INFO1_MASK,
+ INFO1_MASK_CLEAR);
+ renesas_sdhi_internal_dmac_dm_write(host, DM_CM_INFO2_MASK,
+ INFO2_MASK_CLEAR);
+
/* Each value is set to non-zero to assume "enabling" each DMA */
host->chan_rx = host->chan_tx = (void *)0xdeadbeaf;
