acct: fix accuracy loss for input value of encode_comp_t()

Patch series "Fix encode_comp_t()".

Type conversion in encode_comp_t() may look a bit problematic.


This patch (of 2):

See calculation of ac_{u,s}time in fill_ac():
  > ac->ac_utime = encode_comp_t(nsec_to_AHZ(pacct->ac_utime));
  > ac->ac_stime = encode_comp_t(nsec_to_AHZ(pacct->ac_stime));

Return value of nsec_to_AHZ() is always type of 'u64', but it is handled
as type of 'unsigned long' in encode_comp_t, and accuracy loss would
happen on 32-bit platform when 'unsigned long' value is 32-bit-width.

So 'u64' value of encode_comp_t() may look better.

Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Zheng Yejian <[email protected]>
Cc: Hanjun Guo <[email protected]>
Cc: Randy Dunlap <[email protected]> # build-tested
Cc: Vlastimil Babka <[email protected]>
Cc: Zhang Jinhao <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>

kernel/acct.c

@@ -320,7 +320,7 @@ void acct_exit_ns(struct pid_namespace *ns)
 }
 
 /*
- *  encode an unsigned long into a comp_t
+ *  encode an u64 into a comp_t
  *
  *  This routine has been adopted from the encode_comp_t() function in
  *  the kern_acct.c file of the FreeBSD operating system. The encoding
@@ -331,7 +331,7 @@ void acct_exit_ns(struct pid_namespace *ns)
 #define	EXPSIZE		3			/* Base 8 (3 bit) exponent. */
 #define	MAXFRACT	((1 << MANTSIZE) - 1)	/* Maximum fractional value. */
 
-static comp_t encode_comp_t(unsigned long value)
+static comp_t encode_comp_t(u64 value)
 {
 	int exp, rnd;