d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

i386 turns out to not have __uint128_t. So confusingly use 64-bit math

Browse Source 
instead. Since we're little endian, we can get away with it. Also,
since the counters in quesitons would require billions of iops for
tens of billions of seconds to overflow, and since such data rates are
unlikely for people using i386 for a while, that's OK. The fastest
cards today can't do even a million IOPs.

Noticed by: dim@
Sponsored by: Netflix, Inc
This commit is contained in:
Warner Losh 2016-11-19 21:46:13 +00:00
parent 29aee14890
commit 4e44c386ff
Notes: svn2git 2020-12-20 02:59:44 +00:00 
svn path=/head/; revision=308869
1 changed files with 9 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
sbin/nvmecontrol/logpage.c
View File

@ -75,10 +75,18 @@ kv_lookup(const struct kv_name *kv, size_t kv_count, uint32_t key)
} }
/* /*
* 128-bit integer augments to standard values * 128-bit integer augments to standard values. On i386 this
* doesn't exist, so we use 64-bit values. The 128-bit counters
* are crazy anyway, since for this purpose, you'd need a
* billion IOPs for billions of seconds to overflow them.
* So, on 32-bit i386, you'll get truncated values.
*/ */
#define UINT128_DIG 39 #define UINT128_DIG 39
#ifdef __i386__
typedef uint64_t uint128_t;
#else
typedef __uint128_t uint128_t; typedef __uint128_t uint128_t;
#endif
static inline uint128_t static inline uint128_t
to128(void *p) to128(void *p)