2014-12-17 04:24:32

by Chris Rorvick

[permalink] [raw]
Subject: [PATCH v2 0/2] lprocfs_write_frac_u64_helper cleanup

Added a second patch to address Dan Carpenter's concern with the
complexity of passing the sign through `mult'. Compile tested only.

Chris Rorvick (2):
drivers: staging: lustre: Use mult if units not specified
drivers: staging: lustre: Track sign separately

drivers/staging/lustre/lustre/obdclass/lprocfs_status.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)

--
2.1.0


2014-12-17 04:24:34

by Chris Rorvick

[permalink] [raw]
Subject: [PATCH v2 1/2] drivers: staging: lustre: Use mult if units not specified

Units can be passed to lprocfs_write_frac_u64_helper() via a suffix
(e.g., "...K", "...M", etc.) tacked onto the value. A comment states
that "specified units override the multiplier," though the multiplier is
overridden regardless. Update the conditional logic so that it only
applies when units are specified.

Signed-off-by: Chris Rorvick <[email protected]>
---
drivers/staging/lustre/lustre/obdclass/lprocfs_status.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c b/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
index 61e04af..92ed0a0 100644
--- a/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
+++ b/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
@@ -1910,7 +1910,7 @@ int lprocfs_write_frac_u64_helper(const char *buffer, unsigned long count,
units <<= 10;
}
/* Specified units override the multiplier */
- if (units)
+ if (units > 1)
mult = mult < 0 ? -units : units;

frac *= mult;
--
2.1.0

2014-12-17 04:24:49

by Chris Rorvick

[permalink] [raw]
Subject: [PATCH v2 2/2] drivers: staging: lustre: Track sign separately

The `mult' parameter is negated if the user data begins with a '-' so
that the final value has the appropriate sign. But `mult' is only used
if the user data does not include a "units" suffix. In this case,
`mult' is overridden with the numeric scale conveyed by the units suffix,
but retains the sign of the original value.

Having `mult' serving double-duty works but is confusing. Use a new
local variable to store the sign of the user data instead. This also
fixes a pitfall of passing 0 to `mult', expecting it to be ignored when
a units suffix is specified, but having the effect of taking the
absolute value of the user-provided data.

Signed-off-by: Chris Rorvick <[email protected]>
---
drivers/staging/lustre/lustre/obdclass/lprocfs_status.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c b/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
index 92ed0a0..b6c3352 100644
--- a/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
+++ b/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
@@ -1864,6 +1864,7 @@ int lprocfs_write_frac_u64_helper(const char *buffer, unsigned long count,
char kernbuf[22], *end, *pbuf;
__u64 whole, frac = 0, units;
unsigned frac_d = 1;
+ int sign = 1;

if (count > (sizeof(kernbuf) - 1))
return -EINVAL;
@@ -1874,7 +1875,7 @@ int lprocfs_write_frac_u64_helper(const char *buffer, unsigned long count,
kernbuf[count] = '\0';
pbuf = kernbuf;
if (*pbuf == '-') {
- mult = -mult;
+ sign = -1;
pbuf++;
}

@@ -1911,11 +1912,11 @@ int lprocfs_write_frac_u64_helper(const char *buffer, unsigned long count,
}
/* Specified units override the multiplier */
if (units > 1)
- mult = mult < 0 ? -units : units;
+ mult = units;

frac *= mult;
do_div(frac, frac_d);
- *val = whole * mult + frac;
+ *val = sign * (whole * mult + frac);
return 0;
}
EXPORT_SYMBOL(lprocfs_write_frac_u64_helper);
--
2.1.0

2014-12-17 06:54:11

by Dilger, Andreas

[permalink] [raw]
Subject: Re: [PATCH v2 2/2] drivers: staging: lustre: Track sign separately

On 2014/12/16, 9:24 PM, "Chris Rorvick" <[email protected]> wrote:

>The `mult' parameter is negated if the user data begins with a '-' so
>that the final value has the appropriate sign. But `mult' is only used
>if the user data does not include a "units" suffix. In this case,
>`mult' is overridden with the numeric scale conveyed by the units suffix,
>but retains the sign of the original value.
>
>Having `mult' serving double-duty works but is confusing. Use a new
>local variable to store the sign of the user data instead. This also
>fixes a pitfall of passing 0 to `mult', expecting it to be ignored when
>a units suffix is specified, but having the effect of taking the
>absolute value of the user-provided data.
>
>Signed-off-by: Chris Rorvick <[email protected]>

Both patches look good to me.
Reviewed-by: Andreas Dilger <[email protected]>

>---
> drivers/staging/lustre/lustre/obdclass/lprocfs_status.c | 7 ++++---
> 1 file changed, 4 insertions(+), 3 deletions(-)
>
>diff --git a/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
>b/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
>index 92ed0a0..b6c3352 100644
>--- a/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
>+++ b/drivers/staging/lustre/lustre/obdclass/lprocfs_status.c
>@@ -1864,6 +1864,7 @@ int lprocfs_write_frac_u64_helper(const char
>*buffer, unsigned long count,
> char kernbuf[22], *end, *pbuf;
> __u64 whole, frac = 0, units;
> unsigned frac_d = 1;
>+ int sign = 1;
>
> if (count > (sizeof(kernbuf) - 1))
> return -EINVAL;
>@@ -1874,7 +1875,7 @@ int lprocfs_write_frac_u64_helper(const char
>*buffer, unsigned long count,
> kernbuf[count] = '\0';
> pbuf = kernbuf;
> if (*pbuf == '-') {
>- mult = -mult;
>+ sign = -1;
> pbuf++;
> }
>
>@@ -1911,11 +1912,11 @@ int lprocfs_write_frac_u64_helper(const char
>*buffer, unsigned long count,
> }
> /* Specified units override the multiplier */
> if (units > 1)
>- mult = mult < 0 ? -units : units;
>+ mult = units;
>
> frac *= mult;
> do_div(frac, frac_d);
>- *val = whole * mult + frac;
>+ *val = sign * (whole * mult + frac);
> return 0;
> }
> EXPORT_SYMBOL(lprocfs_write_frac_u64_helper);
>--
>2.1.0
>
>


Cheers, Andreas
--
Andreas Dilger

Lustre Software Architect
Intel High Performance Data Division

2014-12-17 07:56:59

by Dan Carpenter

[permalink] [raw]
Subject: Re: [PATCH v2 0/2] lprocfs_write_frac_u64_helper cleanup

On Tue, Dec 16, 2014 at 10:24:00PM -0600, Chris Rorvick wrote:
> Added a second patch to address Dan Carpenter's concern with the
> complexity of passing the sign through `mult'. Compile tested only.


Thanks. That does look cleaner.

regards,
dan carpenter