linux-kernel - Re: [PATCH] Provide u64 version of jiffies_to

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 02 Jan 2008 16:36:47 -0800
From:	"H. Peter Anvin" <hpa@...or.com>
To:	Andrew Morton <akpm@...ux-foundation.org>
CC:	Jonathan Lim <jlim@....com>, linux-kernel@...r.kernel.org,
	jlan@....com
Subject: Re: [PATCH] Provide u64 version of jiffies_to_usecs() in kernel/tsacct.c

Andrew Morton wrote:
> On Fri, 28 Dec 2007 13:26:07 -0800 (PST) Jonathan Lim <jlim@....com> wrote:
> 
>> It's possible that the values used in and returned from jiffies_to_usecs() are
>> incorrect because of truncation when variables of type u64 are involved.  So a
>> function specific to that type is used instead.
>>
>> Diff'd against: linux/kernel/git/stable/linux-2.6.23.y.git
>>
>> Signed-off-by: Jonathan Lim <jlim@....com>
>>
>> --- a/kernel/tsacct.c	2007-12-28 11:58:05.182065029 -0800
>> +++ b/kernel/tsacct.c	2007-12-28 11:57:37.949013675 -0800
>> @@ -71,6 +71,17 @@ void bacct_add_tsk(struct taskstats *sta
>>  
>>  #ifdef CONFIG_TASK_XACCT
>>  
>> +static inline u64 jiffies_to_usecs_u64(const u64 j)
>> +{
>> +#if HZ <= USEC_PER_SEC && !(USEC_PER_SEC % HZ)
>> +	return (USEC_PER_SEC / HZ) * j;
>> +#elif HZ > USEC_PER_SEC && !(HZ % USEC_PER_SEC)
>> +	return (j + (HZ / USEC_PER_SEC) - 1)/(HZ / USEC_PER_SEC);
>> +#else
>> +	return (j * USEC_PER_SEC) / HZ;
>> +#endif
>> +}
>> +
>>  #define KB 1024
>>  #define MB (1024*KB)
>>  /*
>> @@ -81,8 +92,8 @@ void xacct_add_tsk(struct taskstats *sta
>>  	struct mm_struct *mm;
>>  
>>  	/* convert pages-jiffies to Mbyte-usec */
>> -	stats->coremem = jiffies_to_usecs(p->acct_rss_mem1) * PAGE_SIZE / MB;
>> -	stats->virtmem = jiffies_to_usecs(p->acct_vm_mem1) * PAGE_SIZE / MB;
>> +	stats->coremem = jiffies_to_usecs_u64(p->acct_rss_mem1) * PAGE_SIZE / MB;
>> +	stats->virtmem = jiffies_to_usecs_u64(p->acct_vm_mem1) * PAGE_SIZE / MB;
>>  	mm = get_task_mm(p);
>>  	if (mm) {
>>  		/* adjust to KB unit */
> 
> Fair enough.  But I guess that new function should be a kernel-wide thing
> because surely other users will turn up.
> 
> Peter has been working on the accuracy of some of these conversion
> functions and might need to know about this change?

Yes, the function should be coded using the new #defines produced by 
timeconst.h; that way you end up avoiding a possible overflow in the 
multiplication.

I believe all three cases can be folded, then, to:

	return (j*HZ_TO_USEC_NUM + HZ_TO_USEC_DEN-1) / HZ_TO_USEC_DEN;

I would also like to observe that the roundoff behaviour of the function 
  above is inconsistent; in case 2 it will round up, but in case 3 it 
will round down.  The line proposed above has round up behaviour.

	-hpa
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/