Documentation: improvement to volatile considered harmful (resubmit)
PATCH 1/1 Documentation: improvement to volatile considered harmful (resubmit)
I'm resubmitting this as I didn't get any replies, this time CCeing proper people, sorry..
Kernel locking/synchronization primitives are better than volatile types from code readability point of view also.
This patch is against 2.6.22-rc6.
Documentation/volatile-considered-harmful.txt +3 -2
@@ -17,8 +17,9 @@ all optimization-related problems in a more efficient way. Like volatile, the kernel primitives which make concurrent access to data safe (spinlocks, mutexes, memory barriers, etc.) are designed to prevent -unwanted optimization. If they are being used properly, there will be no -need to use volatile as well. If volatile is still necessary, there is +unwanted optimization. If they are being used properly, there will be no +need to use volatile as well. Also, they make code more readable as they +represent their intent explicitly. If volatile is still necessary, there is almost certainly a bug in the code somewhere. In properly-written kernel code, volatile can only serve to slow things down.
Signature · 3 lines
Heikki Orsila Barbie's law: [email protected] "Math is hard, let's go shopping!" http://www.iki.fi/shd
Jonathan Corbet
maintainerQuoted 5 lines · Heikki Orsila
I'm resubmitting this as I didn't get any replies, this time CCeing proper people, sorry..
Kernel locking/synchronization primitives are better than volatile types from code readability point of view also.
I think that just dilutes the real point. It's not a choice between locking and volatile - the locking must be there regardless. It's a correctness issue; if the result happens to be more readable too that's a bonus.
If somebody wants to put this sentence in I won't object, but I don't think it really improves the document either.
Thanks,
jon