* [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
@ 2015-01-23 15:41 ` Maxime Ripard
0 siblings, 0 replies; 15+ messages in thread
From: Maxime Ripard @ 2015-01-23 15:41 UTC (permalink / raw)
To: Gregory Clement, Jason Cooper, Andrew Lunn, Sebastian Hesselbarth,
Ezequiel Garcia, Brian Norris
Cc: linux-mtd, Boris Brezillon, Thomas Petazzoni, linux-arm-kernel,
Tawfik Bayouk, Nadav Haklai, Lior Amsalem, linux-kernel,
Sudhakar Gundubogula, Seif Mazareeb, Maxime Ripard, stable
The NDDB register holds the data that are needed by the read and write
commands.
However, during a read PIO access, the datasheet specifies that after each 32
bits read in that register, when BCH is enabled, we have to make sure that the
RDDREQ bit is set in the NDSR register.
This fixes an issue that was seen on the Armada 385, and presumably other mvebu
SoCs, when a read on a newly erased page would end up in the driver reporting a
timeout from the NAND.
Cc: <stable@vger.kernel.org>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
1 file changed, 30 insertions(+), 6 deletions(-)
diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
index 96b0b1d27df1..320c2ab14d4e 100644
--- a/drivers/mtd/nand/pxa3xx_nand.c
+++ b/drivers/mtd/nand/pxa3xx_nand.c
@@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
nand_writel(info, NDCR, ndcr | int_mask);
}
+static void drain_fifo(struct pxa3xx_nand_info *info,
+ void *data,
+ int len)
+{
+ u32 *dst = (u32 *)data;
+
+ if (info->ecc_bch) {
+ while (len--) {
+ *dst++ = nand_readl(info, NDDB);
+
+ /*
+ * According to the datasheet, when reading
+ * from NDDB with BCH enabled, after each 32
+ * bits reads, we have to make sure that the
+ * NDSR.RDDREQ bit is set
+ */
+ while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
+ cpu_relax();
+ }
+ } else {
+ __raw_readsl(info->mmio_base + NDDB, data, len);
+ }
+}
+
static void handle_data_pio(struct pxa3xx_nand_info *info)
{
unsigned int do_bytes = min(info->data_size, info->chunk_size);
@@ -496,14 +520,14 @@ static void handle_data_pio(struct pxa3xx_nand_info *info)
DIV_ROUND_UP(info->oob_size, 4));
break;
case STATE_PIO_READING:
- __raw_readsl(info->mmio_base + NDDB,
- info->data_buff + info->data_buff_pos,
- DIV_ROUND_UP(do_bytes, 4));
+ drain_fifo(info,
+ info->data_buff + info->data_buff_pos,
+ DIV_ROUND_UP(do_bytes, 4));
if (info->oob_size > 0)
- __raw_readsl(info->mmio_base + NDDB,
- info->oob_buff + info->oob_buff_pos,
- DIV_ROUND_UP(info->oob_size, 4));
+ drain_fifo(info,
+ info->oob_buff + info->oob_buff_pos,
+ DIV_ROUND_UP(info->oob_size, 4));
break;
default:
dev_err(&info->pdev->dev, "%s: invalid state %d\n", __func__,
--
2.2.2
^ permalink raw reply related [flat|nested] 15+ messages in thread* [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
@ 2015-01-23 15:41 ` Maxime Ripard
0 siblings, 0 replies; 15+ messages in thread
From: Maxime Ripard @ 2015-01-23 15:41 UTC (permalink / raw)
To: linux-arm-kernel
The NDDB register holds the data that are needed by the read and write
commands.
However, during a read PIO access, the datasheet specifies that after each 32
bits read in that register, when BCH is enabled, we have to make sure that the
RDDREQ bit is set in the NDSR register.
This fixes an issue that was seen on the Armada 385, and presumably other mvebu
SoCs, when a read on a newly erased page would end up in the driver reporting a
timeout from the NAND.
Cc: <stable@vger.kernel.org>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
1 file changed, 30 insertions(+), 6 deletions(-)
diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
index 96b0b1d27df1..320c2ab14d4e 100644
--- a/drivers/mtd/nand/pxa3xx_nand.c
+++ b/drivers/mtd/nand/pxa3xx_nand.c
@@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
nand_writel(info, NDCR, ndcr | int_mask);
}
+static void drain_fifo(struct pxa3xx_nand_info *info,
+ void *data,
+ int len)
+{
+ u32 *dst = (u32 *)data;
+
+ if (info->ecc_bch) {
+ while (len--) {
+ *dst++ = nand_readl(info, NDDB);
+
+ /*
+ * According to the datasheet, when reading
+ * from NDDB with BCH enabled, after each 32
+ * bits reads, we have to make sure that the
+ * NDSR.RDDREQ bit is set
+ */
+ while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
+ cpu_relax();
+ }
+ } else {
+ __raw_readsl(info->mmio_base + NDDB, data, len);
+ }
+}
+
static void handle_data_pio(struct pxa3xx_nand_info *info)
{
unsigned int do_bytes = min(info->data_size, info->chunk_size);
@@ -496,14 +520,14 @@ static void handle_data_pio(struct pxa3xx_nand_info *info)
DIV_ROUND_UP(info->oob_size, 4));
break;
case STATE_PIO_READING:
- __raw_readsl(info->mmio_base + NDDB,
- info->data_buff + info->data_buff_pos,
- DIV_ROUND_UP(do_bytes, 4));
+ drain_fifo(info,
+ info->data_buff + info->data_buff_pos,
+ DIV_ROUND_UP(do_bytes, 4));
if (info->oob_size > 0)
- __raw_readsl(info->mmio_base + NDDB,
- info->oob_buff + info->oob_buff_pos,
- DIV_ROUND_UP(info->oob_size, 4));
+ drain_fifo(info,
+ info->oob_buff + info->oob_buff_pos,
+ DIV_ROUND_UP(info->oob_size, 4));
break;
default:
dev_err(&info->pdev->dev, "%s: invalid state %d\n", __func__,
--
2.2.2
^ permalink raw reply related [flat|nested] 15+ messages in thread* Re: [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
2015-01-23 15:41 ` Maxime Ripard
(?)
@ 2015-01-23 15:59 ` Gregory CLEMENT
-1 siblings, 0 replies; 15+ messages in thread
From: Gregory CLEMENT @ 2015-01-23 15:59 UTC (permalink / raw)
To: Maxime Ripard
Cc: Lior Amsalem, Andrew Lunn, Jason Cooper, Tawfik Bayouk,
Thomas Petazzoni, Seif Mazareeb, linux-kernel, stable,
Sudhakar Gundubogula, Nadav Haklai, Boris Brezillon, linux-mtd,
Ezequiel Garcia, Brian Norris, linux-arm-kernel,
Sebastian Hesselbarth
Hi Maxime,
On 23/01/2015 16:41, Maxime Ripard wrote:
> The NDDB register holds the data that are needed by the read and write
> commands.
>
> However, during a read PIO access, the datasheet specifies that after each 32
> bits read in that register, when BCH is enabled, we have to make sure that the
> RDDREQ bit is set in the NDSR register.
>
> This fixes an issue that was seen on the Armada 385, and presumably other mvebu
> SoCs, when a read on a newly erased page would end up in the driver reporting a
> timeout from the NAND.
>
> Cc: <stable@vger.kernel.org>
It would help the stable maintainer if you could indicate since which commit or
kernel release this fix should be applied.
> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
> ---
> drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
> 1 file changed, 30 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
> index 96b0b1d27df1..320c2ab14d4e 100644
> --- a/drivers/mtd/nand/pxa3xx_nand.c
> +++ b/drivers/mtd/nand/pxa3xx_nand.c
> @@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
> nand_writel(info, NDCR, ndcr | int_mask);
> }
>
> +static void drain_fifo(struct pxa3xx_nand_info *info,
> + void *data,
> + int len)
> +{
> + u32 *dst = (u32 *)data;
> +
> + if (info->ecc_bch) {
> + while (len--) {
> + *dst++ = nand_readl(info, NDDB);
> +
> + /*
> + * According to the datasheet, when reading
> + * from NDDB with BCH enabled, after each 32
> + * bits reads, we have to make sure that the
> + * NDSR.RDDREQ bit is set
> + */
> + while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
> + cpu_relax();
Are we sure that we won't be blocked here?
If not, what about adding a timeout?
Thanks,
Gregory
> + }
> + } else {
> + __raw_readsl(info->mmio_base + NDDB, data, len);
> + }
> +}
> +
> static void handle_data_pio(struct pxa3xx_nand_info *info)
> {
> unsigned int do_bytes = min(info->data_size, info->chunk_size);
> @@ -496,14 +520,14 @@ static void handle_data_pio(struct pxa3xx_nand_info *info)
> DIV_ROUND_UP(info->oob_size, 4));
> break;
> case STATE_PIO_READING:
> - __raw_readsl(info->mmio_base + NDDB,
> - info->data_buff + info->data_buff_pos,
> - DIV_ROUND_UP(do_bytes, 4));
> + drain_fifo(info,
> + info->data_buff + info->data_buff_pos,
> + DIV_ROUND_UP(do_bytes, 4));
>
> if (info->oob_size > 0)
> - __raw_readsl(info->mmio_base + NDDB,
> - info->oob_buff + info->oob_buff_pos,
> - DIV_ROUND_UP(info->oob_size, 4));
> + drain_fifo(info,
> + info->oob_buff + info->oob_buff_pos,
> + DIV_ROUND_UP(info->oob_size, 4));
> break;
> default:
> dev_err(&info->pdev->dev, "%s: invalid state %d\n", __func__,
>
--
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
^ permalink raw reply [flat|nested] 15+ messages in thread* Re: [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
@ 2015-01-23 15:59 ` Gregory CLEMENT
0 siblings, 0 replies; 15+ messages in thread
From: Gregory CLEMENT @ 2015-01-23 15:59 UTC (permalink / raw)
To: Maxime Ripard
Cc: Jason Cooper, Andrew Lunn, Sebastian Hesselbarth, Ezequiel Garcia,
Brian Norris, linux-mtd, Boris Brezillon, Thomas Petazzoni,
linux-arm-kernel, Tawfik Bayouk, Nadav Haklai, Lior Amsalem,
linux-kernel, Sudhakar Gundubogula, Seif Mazareeb, stable
Hi Maxime,
On 23/01/2015 16:41, Maxime Ripard wrote:
> The NDDB register holds the data that are needed by the read and write
> commands.
>
> However, during a read PIO access, the datasheet specifies that after each 32
> bits read in that register, when BCH is enabled, we have to make sure that the
> RDDREQ bit is set in the NDSR register.
>
> This fixes an issue that was seen on the Armada 385, and presumably other mvebu
> SoCs, when a read on a newly erased page would end up in the driver reporting a
> timeout from the NAND.
>
> Cc: <stable@vger.kernel.org>
It would help the stable maintainer if you could indicate since which commit or
kernel release this fix should be applied.
> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
> ---
> drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
> 1 file changed, 30 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
> index 96b0b1d27df1..320c2ab14d4e 100644
> --- a/drivers/mtd/nand/pxa3xx_nand.c
> +++ b/drivers/mtd/nand/pxa3xx_nand.c
> @@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
> nand_writel(info, NDCR, ndcr | int_mask);
> }
>
> +static void drain_fifo(struct pxa3xx_nand_info *info,
> + void *data,
> + int len)
> +{
> + u32 *dst = (u32 *)data;
> +
> + if (info->ecc_bch) {
> + while (len--) {
> + *dst++ = nand_readl(info, NDDB);
> +
> + /*
> + * According to the datasheet, when reading
> + * from NDDB with BCH enabled, after each 32
> + * bits reads, we have to make sure that the
> + * NDSR.RDDREQ bit is set
> + */
> + while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
> + cpu_relax();
Are we sure that we won't be blocked here?
If not, what about adding a timeout?
Thanks,
Gregory
> + }
> + } else {
> + __raw_readsl(info->mmio_base + NDDB, data, len);
> + }
> +}
> +
> static void handle_data_pio(struct pxa3xx_nand_info *info)
> {
> unsigned int do_bytes = min(info->data_size, info->chunk_size);
> @@ -496,14 +520,14 @@ static void handle_data_pio(struct pxa3xx_nand_info *info)
> DIV_ROUND_UP(info->oob_size, 4));
> break;
> case STATE_PIO_READING:
> - __raw_readsl(info->mmio_base + NDDB,
> - info->data_buff + info->data_buff_pos,
> - DIV_ROUND_UP(do_bytes, 4));
> + drain_fifo(info,
> + info->data_buff + info->data_buff_pos,
> + DIV_ROUND_UP(do_bytes, 4));
>
> if (info->oob_size > 0)
> - __raw_readsl(info->mmio_base + NDDB,
> - info->oob_buff + info->oob_buff_pos,
> - DIV_ROUND_UP(info->oob_size, 4));
> + drain_fifo(info,
> + info->oob_buff + info->oob_buff_pos,
> + DIV_ROUND_UP(info->oob_size, 4));
> break;
> default:
> dev_err(&info->pdev->dev, "%s: invalid state %d\n", __func__,
>
--
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
^ permalink raw reply [flat|nested] 15+ messages in thread* [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
@ 2015-01-23 15:59 ` Gregory CLEMENT
0 siblings, 0 replies; 15+ messages in thread
From: Gregory CLEMENT @ 2015-01-23 15:59 UTC (permalink / raw)
To: linux-arm-kernel
Hi Maxime,
On 23/01/2015 16:41, Maxime Ripard wrote:
> The NDDB register holds the data that are needed by the read and write
> commands.
>
> However, during a read PIO access, the datasheet specifies that after each 32
> bits read in that register, when BCH is enabled, we have to make sure that the
> RDDREQ bit is set in the NDSR register.
>
> This fixes an issue that was seen on the Armada 385, and presumably other mvebu
> SoCs, when a read on a newly erased page would end up in the driver reporting a
> timeout from the NAND.
>
> Cc: <stable@vger.kernel.org>
It would help the stable maintainer if you could indicate since which commit or
kernel release this fix should be applied.
> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
> ---
> drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
> 1 file changed, 30 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
> index 96b0b1d27df1..320c2ab14d4e 100644
> --- a/drivers/mtd/nand/pxa3xx_nand.c
> +++ b/drivers/mtd/nand/pxa3xx_nand.c
> @@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
> nand_writel(info, NDCR, ndcr | int_mask);
> }
>
> +static void drain_fifo(struct pxa3xx_nand_info *info,
> + void *data,
> + int len)
> +{
> + u32 *dst = (u32 *)data;
> +
> + if (info->ecc_bch) {
> + while (len--) {
> + *dst++ = nand_readl(info, NDDB);
> +
> + /*
> + * According to the datasheet, when reading
> + * from NDDB with BCH enabled, after each 32
> + * bits reads, we have to make sure that the
> + * NDSR.RDDREQ bit is set
> + */
> + while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
> + cpu_relax();
Are we sure that we won't be blocked here?
If not, what about adding a timeout?
Thanks,
Gregory
> + }
> + } else {
> + __raw_readsl(info->mmio_base + NDDB, data, len);
> + }
> +}
> +
> static void handle_data_pio(struct pxa3xx_nand_info *info)
> {
> unsigned int do_bytes = min(info->data_size, info->chunk_size);
> @@ -496,14 +520,14 @@ static void handle_data_pio(struct pxa3xx_nand_info *info)
> DIV_ROUND_UP(info->oob_size, 4));
> break;
> case STATE_PIO_READING:
> - __raw_readsl(info->mmio_base + NDDB,
> - info->data_buff + info->data_buff_pos,
> - DIV_ROUND_UP(do_bytes, 4));
> + drain_fifo(info,
> + info->data_buff + info->data_buff_pos,
> + DIV_ROUND_UP(do_bytes, 4));
>
> if (info->oob_size > 0)
> - __raw_readsl(info->mmio_base + NDDB,
> - info->oob_buff + info->oob_buff_pos,
> - DIV_ROUND_UP(info->oob_size, 4));
> + drain_fifo(info,
> + info->oob_buff + info->oob_buff_pos,
> + DIV_ROUND_UP(info->oob_size, 4));
> break;
> default:
> dev_err(&info->pdev->dev, "%s: invalid state %d\n", __func__,
>
--
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
^ permalink raw reply [flat|nested] 15+ messages in thread* Re: [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
2015-01-23 15:59 ` Gregory CLEMENT
(?)
@ 2015-01-24 14:01 ` Ezequiel Garcia
-1 siblings, 0 replies; 15+ messages in thread
From: Ezequiel Garcia @ 2015-01-24 14:01 UTC (permalink / raw)
To: Gregory CLEMENT, Maxime Ripard
Cc: Lior Amsalem, Andrew Lunn, Jason Cooper, Tawfik Bayouk,
Thomas Petazzoni, Seif Mazareeb, linux-kernel, stable,
Sudhakar Gundubogula, Nadav Haklai, Boris Brezillon, linux-mtd,
Brian Norris, linux-arm-kernel, Sebastian Hesselbarth
On 01/23/2015 12:59 PM, Gregory CLEMENT wrote:
> On 23/01/2015 16:41, Maxime Ripard wrote:
>> The NDDB register holds the data that are needed by the read and write
>> commands.
>>
>> However, during a read PIO access, the datasheet specifies that after each 32
>> bits read in that register, when BCH is enabled, we have to make sure that the
>> RDDREQ bit is set in the NDSR register.
>>
>> This fixes an issue that was seen on the Armada 385, and presumably other mvebu
>> SoCs, when a read on a newly erased page would end up in the driver reporting a
>> timeout from the NAND.
>>
>> Cc: <stable@vger.kernel.org>
>
> It would help the stable maintainer if you could indicate since which commit or
> kernel release this fix should be applied.
>
This is a fix for the BCH support, namely commit 43bcfd2bb24a
"mtd: nand: pxa3xx: Add driver-specific ECC BCH support". The commit was merged
in v3.14.
However, this patch won't apply directly there. It will apply on commit
fa543bef72d6 "mtd: nand: pxa3xx: Add a read/write buffers markers"; which
was also merged in v3.14.
Therefore, I guess it's OK to say
Cc: <stable@vger.kernel.org> # v3.14.x
>> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
>> ---
>> drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
>> 1 file changed, 30 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
>> index 96b0b1d27df1..320c2ab14d4e 100644
>> --- a/drivers/mtd/nand/pxa3xx_nand.c
>> +++ b/drivers/mtd/nand/pxa3xx_nand.c
>> @@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
>> nand_writel(info, NDCR, ndcr | int_mask);
>> }
>>
>> +static void drain_fifo(struct pxa3xx_nand_info *info,
>> + void *data,
>> + int len)
^^
You don't need to split that line, it seems to fit 80 characters as is.
>> +{
>> + u32 *dst = (u32 *)data;
>> +
>> + if (info->ecc_bch) {
>> + while (len--) {
>> + *dst++ = nand_readl(info, NDDB);
>> +
>> + /*
>> + * According to the datasheet, when reading
>> + * from NDDB with BCH enabled, after each 32
>> + * bits reads, we have to make sure that the
>> + * NDSR.RDDREQ bit is set
>> + */
>> + while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
>> + cpu_relax();
>
> Are we sure that we won't be blocked here?
> If not, what about adding a timeout?
>
Definitely. I think we shouldn't have an infinite loop, no matter what the hw specs say.
--
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 15+ messages in thread* Re: [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
@ 2015-01-24 14:01 ` Ezequiel Garcia
0 siblings, 0 replies; 15+ messages in thread
From: Ezequiel Garcia @ 2015-01-24 14:01 UTC (permalink / raw)
To: Gregory CLEMENT, Maxime Ripard
Cc: Jason Cooper, Andrew Lunn, Sebastian Hesselbarth, Brian Norris,
linux-mtd, Boris Brezillon, Thomas Petazzoni, linux-arm-kernel,
Tawfik Bayouk, Nadav Haklai, Lior Amsalem, linux-kernel,
Sudhakar Gundubogula, Seif Mazareeb, stable
On 01/23/2015 12:59 PM, Gregory CLEMENT wrote:
> On 23/01/2015 16:41, Maxime Ripard wrote:
>> The NDDB register holds the data that are needed by the read and write
>> commands.
>>
>> However, during a read PIO access, the datasheet specifies that after each 32
>> bits read in that register, when BCH is enabled, we have to make sure that the
>> RDDREQ bit is set in the NDSR register.
>>
>> This fixes an issue that was seen on the Armada 385, and presumably other mvebu
>> SoCs, when a read on a newly erased page would end up in the driver reporting a
>> timeout from the NAND.
>>
>> Cc: <stable@vger.kernel.org>
>
> It would help the stable maintainer if you could indicate since which commit or
> kernel release this fix should be applied.
>
This is a fix for the BCH support, namely commit 43bcfd2bb24a
"mtd: nand: pxa3xx: Add driver-specific ECC BCH support". The commit was merged
in v3.14.
However, this patch won't apply directly there. It will apply on commit
fa543bef72d6 "mtd: nand: pxa3xx: Add a read/write buffers markers"; which
was also merged in v3.14.
Therefore, I guess it's OK to say
Cc: <stable@vger.kernel.org> # v3.14.x
>> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
>> ---
>> drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
>> 1 file changed, 30 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
>> index 96b0b1d27df1..320c2ab14d4e 100644
>> --- a/drivers/mtd/nand/pxa3xx_nand.c
>> +++ b/drivers/mtd/nand/pxa3xx_nand.c
>> @@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
>> nand_writel(info, NDCR, ndcr | int_mask);
>> }
>>
>> +static void drain_fifo(struct pxa3xx_nand_info *info,
>> + void *data,
>> + int len)
^^
You don't need to split that line, it seems to fit 80 characters as is.
>> +{
>> + u32 *dst = (u32 *)data;
>> +
>> + if (info->ecc_bch) {
>> + while (len--) {
>> + *dst++ = nand_readl(info, NDDB);
>> +
>> + /*
>> + * According to the datasheet, when reading
>> + * from NDDB with BCH enabled, after each 32
>> + * bits reads, we have to make sure that the
>> + * NDSR.RDDREQ bit is set
>> + */
>> + while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
>> + cpu_relax();
>
> Are we sure that we won't be blocked here?
> If not, what about adding a timeout?
>
Definitely. I think we shouldn't have an infinite loop, no matter what the hw specs say.
--
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 15+ messages in thread* [PATCH 1/2] mtd: nand: pxa3xx: Fix PIO FIFO draining
@ 2015-01-24 14:01 ` Ezequiel Garcia
0 siblings, 0 replies; 15+ messages in thread
From: Ezequiel Garcia @ 2015-01-24 14:01 UTC (permalink / raw)
To: linux-arm-kernel
On 01/23/2015 12:59 PM, Gregory CLEMENT wrote:
> On 23/01/2015 16:41, Maxime Ripard wrote:
>> The NDDB register holds the data that are needed by the read and write
>> commands.
>>
>> However, during a read PIO access, the datasheet specifies that after each 32
>> bits read in that register, when BCH is enabled, we have to make sure that the
>> RDDREQ bit is set in the NDSR register.
>>
>> This fixes an issue that was seen on the Armada 385, and presumably other mvebu
>> SoCs, when a read on a newly erased page would end up in the driver reporting a
>> timeout from the NAND.
>>
>> Cc: <stable@vger.kernel.org>
>
> It would help the stable maintainer if you could indicate since which commit or
> kernel release this fix should be applied.
>
This is a fix for the BCH support, namely commit 43bcfd2bb24a
"mtd: nand: pxa3xx: Add driver-specific ECC BCH support". The commit was merged
in v3.14.
However, this patch won't apply directly there. It will apply on commit
fa543bef72d6 "mtd: nand: pxa3xx: Add a read/write buffers markers"; which
was also merged in v3.14.
Therefore, I guess it's OK to say
Cc: <stable@vger.kernel.org> # v3.14.x
>> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
>> ---
>> drivers/mtd/nand/pxa3xx_nand.c | 36 ++++++++++++++++++++++++++++++------
>> 1 file changed, 30 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
>> index 96b0b1d27df1..320c2ab14d4e 100644
>> --- a/drivers/mtd/nand/pxa3xx_nand.c
>> +++ b/drivers/mtd/nand/pxa3xx_nand.c
>> @@ -480,6 +480,30 @@ static void disable_int(struct pxa3xx_nand_info *info, uint32_t int_mask)
>> nand_writel(info, NDCR, ndcr | int_mask);
>> }
>>
>> +static void drain_fifo(struct pxa3xx_nand_info *info,
>> + void *data,
>> + int len)
^^
You don't need to split that line, it seems to fit 80 characters as is.
>> +{
>> + u32 *dst = (u32 *)data;
>> +
>> + if (info->ecc_bch) {
>> + while (len--) {
>> + *dst++ = nand_readl(info, NDDB);
>> +
>> + /*
>> + * According to the datasheet, when reading
>> + * from NDDB with BCH enabled, after each 32
>> + * bits reads, we have to make sure that the
>> + * NDSR.RDDREQ bit is set
>> + */
>> + while (!(nand_readl(info, NDSR) & NDSR_RDDREQ))
>> + cpu_relax();
>
> Are we sure that we won't be blocked here?
> If not, what about adding a timeout?
>
Definitely. I think we shouldn't have an infinite loop, no matter what the hw specs say.
--
Ezequiel Garc?a, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 15+ messages in thread