2017-04-04 23:35:17

by Martin K. Petersen

Subject: Re: [PATCH] sd: close hole in > 2T device rejection when !CONFIG_LBDAF

Bart Van Assche <[email protected]> writes:

Hi Bart,

> Sorry but I still don't understand why the two checks are
> different. How about the (untested) patch below? The approach below
> avoids that the check is duplicated and - at least in my opinion -
> results in code that is easier to read.

Just tripped over this issue in connection with something else. However,
I had to make a few passes to convince myself that your proposed fix was
correct. How about something like the following?

Martin

diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index fb9b4d29af0b..6084c415c070 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -2102,6 +2102,16 @@ static void read_capacity_error(struct scsi_disk *sdkp, struct scsi_device *sdp,

#define READ_CAPACITY_RETRIES_ON_RESET 10

+static bool sd_addressable_capacity(u64 lba, unsigned int sector_size)
+{
+ u64 last_sector = lba + 1ULL << ilog2(sector_size) - 9;
+
+ if (sizeof(sector_t) == 4 && last_sector > 0xffffffffULL)
+ return false;
+
+ return true;
+}
+
static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp,
unsigned char *buffer)
{
@@ -2167,7 +2177,7 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp,
return -ENODEV;
}

- if ((sizeof(sdkp->capacity) == 4) && (lba >= 0xffffffffULL)) {
+ if (!sd_addressable_capacity(lba, sector_size)) {
sd_printk(KERN_ERR, sdkp, "Too big for this kernel. Use a "
"kernel compiled with support for large block "
"devices.\n");
@@ -2256,7 +2266,7 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp,
return sector_size;
}

- if ((sizeof(sdkp->capacity) == 4) && (lba == 0xffffffff)) {
+ if (!sd_addressable_capacity(lba, sector_size)) {
sd_printk(KERN_ERR, sdkp, "Too big for this kernel. Use a "
"kernel compiled with support for large block "
"devices.\n");

2017-04-04 23:55:02

by Bart Van Assche

Subject: Re: [PATCH] sd: close hole in > 2T device rejection when !CONFIG_LBDAF

On Tue, 2017-04-04 at 19:35 -0400, Martin K. Petersen wrote:
> diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
> index fb9b4d29af0b..6084c415c070 100644
> --- a/drivers/scsi/sd.c
> +++ b/drivers/scsi/sd.c
> @@ -2102,6 +2102,16 @@ static void read_capacity_error(struct scsi_disk *sdkp, struct scsi_device *sdp,
>
> #define READ_CAPACITY_RETRIES_ON_RESET 10
>
> +static bool sd_addressable_capacity(u64 lba, unsigned int sector_size)
> +{
> + u64 last_sector = lba + 1ULL << ilog2(sector_size) - 9;
> +
> + if (sizeof(sector_t) == 4 && last_sector > 0xffffffffULL)
> + return false;
> +
> + return true;
> +}

Hello Martin,

How about replacing 0xffffffffULL with U32_MAX, adding parentheses in the
last_sector computation to make clear that + and - have precedence over <<
and adding a comment above?sd_addressable_capacity() that explains its
purpose and also that the shift operation must not be replaced with a call
to logical_to_sectors()? Otherwise this patch looks fine to me.

Thanks,

Bart.