[PATCH] tools: fw_env: use erasesize from MEMGETINFO ioctl
We have a board with several revisions. The older ones use a nor flash with 64k erase size, while the newer have a flash with 4k sectors. The environment size is 8k.
Currently, we have to put a column containing 0x10000 (64k) in fw_env.config in order for it to work on the older boards. But that ends up wasting quite a lot of time on the newer boards that could just erase the 8k occupied by the environment - strace says the 64k erase takes 0.405 seconds. With this patch, as expected, that's about an 8-fold better, at 0.043 seconds.
Having different fw_env.config files for the different revisions is highly impractical, and the correct information is already available right at our fingertips. So use the erasesize returned by the MEMGETINFO ioctl when the fourth column is absent or contains a 0.
As I'm only testing this on a NOR flash, I'm only changing the logic for that case, though I think it should be possible for the other types as well.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk --- tools/env/fw_env.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/tools/env/fw_env.c b/tools/env/fw_env.c index 381739d28d..87c3504315 100644 --- a/tools/env/fw_env.c +++ b/tools/env/fw_env.c @@ -1647,6 +1647,8 @@ static int check_device_config(int dev) goto err; } DEVTYPE(dev) = mtdinfo.type; + if (DEVESIZE(dev) == 0 && mtdinfo.type == MTD_NORFLASH) + DEVESIZE(dev) = mtdinfo.erasesize; if (DEVESIZE(dev) == 0) /* Assume the erase size is the same as the env-size */ DEVESIZE(dev) = ENVSIZE(dev);
Dear Rasmus,
In message 20200320140414.19689-1-rasmus.villemoes@prevas.dk you wrote:
Having different fw_env.config files for the different revisions is highly impractical, and the correct information is already available right at our fingertips. So use the erasesize returned by the MEMGETINFO ioctl when the fourth column is absent or contains a 0.
As I'm only testing this on a NOR flash, I'm only changing the logic for that case, though I think it should be possible for the other types as well.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk
tools/env/fw_env.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/tools/env/fw_env.c b/tools/env/fw_env.c index 381739d28d..87c3504315 100644 --- a/tools/env/fw_env.c +++ b/tools/env/fw_env.c @@ -1647,6 +1647,8 @@ static int check_device_config(int dev) goto err; } DEVTYPE(dev) = mtdinfo.type;
if (DEVESIZE(dev) == 0 && mtdinfo.type == MTD_NORFLASH)DEVESIZE(dev) = mtdinfo.erasesize;
What happens if you - say - have an environment size of 16 KiB and an erase block size of 4 KiB?
Best regards,
Wolfgang Denk
On 20/03/2020 15.11, Wolfgang Denk wrote:
Dear Rasmus,
In message 20200320140414.19689-1-rasmus.villemoes@prevas.dk you wrote:
Having different fw_env.config files for the different revisions is highly impractical, and the correct information is already available right at our fingertips. So use the erasesize returned by the MEMGETINFO ioctl when the fourth column is absent or contains a 0.
As I'm only testing this on a NOR flash, I'm only changing the logic for that case, though I think it should be possible for the other types as well.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk
tools/env/fw_env.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/tools/env/fw_env.c b/tools/env/fw_env.c index 381739d28d..87c3504315 100644 --- a/tools/env/fw_env.c +++ b/tools/env/fw_env.c @@ -1647,6 +1647,8 @@ static int check_device_config(int dev) goto err; } DEVTYPE(dev) = mtdinfo.type;
if (DEVESIZE(dev) == 0 && mtdinfo.type == MTD_NORFLASH)DEVESIZE(dev) = mtdinfo.erasesize;What happens if you - say - have an environment size of 16 KiB and an erase block size of 4 KiB?
The right thing - just as in my test case of an environment of 8K and erase block size of 4K, ENVSECTORS gets computed as 2 (or 4 in your example). And if the environment has a size which is not a multiple of the erase block size, all the same logic as usual applies.
It really works just the same as if one explicitly put in an erase block size in fw_env.config of 4K, but as I wrote in the commit log, there are cases where one has to support different flashes with different erase block sizes.
Now, if somebody has put in 0 for the erase block size (relying on the env size being used) and a 1 in the ENVSECTORS column, then yes, this will break. I can certainly add an additional check for ENVSECTORS(dev) being 0 in order to apply the mtdinfo.erasesize.
Rasmus
On 20/03/2020 15.27, Rasmus Villemoes wrote:
On 20/03/2020 15.11, Wolfgang Denk wrote:
Dear Rasmus,
In message 20200320140414.19689-1-rasmus.villemoes@prevas.dk you wrote:
Having different fw_env.config files for the different revisions is highly impractical, and the correct information is already available right at our fingertips. So use the erasesize returned by the MEMGETINFO ioctl when the fourth column is absent or contains a 0.
As I'm only testing this on a NOR flash, I'm only changing the logic for that case, though I think it should be possible for the other types as well.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk
tools/env/fw_env.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/tools/env/fw_env.c b/tools/env/fw_env.c index 381739d28d..87c3504315 100644 --- a/tools/env/fw_env.c +++ b/tools/env/fw_env.c @@ -1647,6 +1647,8 @@ static int check_device_config(int dev) goto err; } DEVTYPE(dev) = mtdinfo.type;
if (DEVESIZE(dev) == 0 && mtdinfo.type == MTD_NORFLASH)DEVESIZE(dev) = mtdinfo.erasesize;What happens if you - say - have an environment size of 16 KiB and an erase block size of 4 KiB?
The right thing - just as in my test case of an environment of 8K and erase block size of 4K, ENVSECTORS gets computed as 2 (or 4 in your example). And if the environment has a size which is not a multiple of the erase block size, all the same logic as usual applies.
It really works just the same as if one explicitly put in an erase block size in fw_env.config of 4K, but as I wrote in the commit log, there are cases where one has to support different flashes with different erase block sizes.
Now, if somebody has put in 0 for the erase block size (relying on the env size being used) and a 1 in the ENVSECTORS column, then yes, this will break.
But, it won't break silently - there's already a check
if (ENVSIZE(dev) > ENVSECTORS(dev) * DEVESIZE(dev)) {
Rasmus
We have a board with several revisions. The older ones use a nor flash with 64k erase size, while the newer have a flash with 4k sectors. The environment size is 8k.
Currently, we have to put a column containing 0x10000 (64k) in fw_env.config in order for it to work on the older boards. But that ends up wasting quite a lot of time on the newer boards that could just erase the 8k occupied by the environment - strace says the 64k erase takes 0.405 seconds. With this patch, as expected, that's about an 8-fold better, at 0.043 seconds.
Having different fw_env.config files for the different revisions is highly impractical, and the correct information is already available right at our fingertips. So use the erasesize returned by the MEMGETINFO ioctl when the fourth and fifth columns (sector size and #sectors, respectively) are absent or contain 0, a case where the logic previously used to use the environment size as erase size (and consequently computed ENVSECTORS(dev) as 1).
As I'm only testing this on a NOR flash, I'm only changing the logic for that case, though I think it should be possible for the other types as well.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk ---
v2: Also require ENVSECTORS(dev) to be absent (or 0) to apply this, in order to prevent regressions in the somewhat odd case where someone used 0 as sector size (and hence used the environment size as sector size) and explicitly listed 1 as #sectors.
tools/env/fw_env.c | 3 +++ 1 file changed, 3 insertions(+)
diff --git a/tools/env/fw_env.c b/tools/env/fw_env.c index 381739d28d..8734663cd4 100644 --- a/tools/env/fw_env.c +++ b/tools/env/fw_env.c @@ -1647,6 +1647,9 @@ static int check_device_config(int dev) goto err; } DEVTYPE(dev) = mtdinfo.type; + if (DEVESIZE(dev) == 0 && ENVSECTORS(dev) == 0 && + mtdinfo.type == MTD_NORFLASH) + DEVESIZE(dev) = mtdinfo.erasesize; if (DEVESIZE(dev) == 0) /* Assume the erase size is the same as the env-size */ DEVESIZE(dev) = ENVSIZE(dev);
On Tue, Mar 24, 2020 at 01:57:40PM +0100, Rasmus Villemoes wrote:
We have a board with several revisions. The older ones use a nor flash with 64k erase size, while the newer have a flash with 4k sectors. The environment size is 8k.
Currently, we have to put a column containing 0x10000 (64k) in fw_env.config in order for it to work on the older boards. But that ends up wasting quite a lot of time on the newer boards that could just erase the 8k occupied by the environment - strace says the 64k erase takes 0.405 seconds. With this patch, as expected, that's about an 8-fold better, at 0.043 seconds.
Having different fw_env.config files for the different revisions is highly impractical, and the correct information is already available right at our fingertips. So use the erasesize returned by the MEMGETINFO ioctl when the fourth and fifth columns (sector size and #sectors, respectively) are absent or contain 0, a case where the logic previously used to use the environment size as erase size (and consequently computed ENVSECTORS(dev) as 1).
As I'm only testing this on a NOR flash, I'm only changing the logic for that case, though I think it should be possible for the other types as well.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk
Applied to u-boot/master, thanks!
participants (3)
-
Rasmus Villemoes -
Tom Rini -
Wolfgang Denk