[acqq]

But is the argument for higher consistency based on the existence of "saner" architecture as in x86, and wasn't argument insufficient in the time the article was written (1998) as well as is now, given that ARM predates the article and its popularity significantly increased since?

[tfoil2]

I see the reference to the "1998 ACM Subject Classification", but this paper is part of a May 2015 conference: http://www.cs.ucla.edu/~todd/research/pub.php?id=snapl15

[avandeursen]

Yes, it is from the "Summit oN Advances in Programming Languages (SNAPL)", May 3-6, 2015.

http://snapl.org/2015/

[acqq]

Then ignoring the existence of the weaker memory model of ARM (versus the stronger x86/x64) is even stranger. And even if x86/x64 models are stronger, they still need some fences and using them all the time would be too slow. So I still don't really understand the arguments of the article.

[madanMus]

The article does not ignore architectures like ARM. You do need fences, but not all the time - the compiler can avoid fences in places where there is no danger of violating sequential consistency (e.g. on accesses to provably local objects).