Subj : Re: Non-strictly-conforming and unspecified versus undefined behavior To : comp.programming.threads,comp.std.c From : David Schwartz Date : Tue Feb 22 2005 01:38 pm "Marcin 'Qrczak' Kowalczyk" wrote in message news:87wtt0i8oo.fsf@qrnik.zagroda... > "David Schwartz" writes: >>> Not "a lot", because thread-specific addresses can be implemented >>> in terms of thread-specific objects with only one extra indirection. >> On every conceivable platform? He said it "can be alot faster on some >> platforms" and you reply with an optimization that works only on some >> platforms (and often burns a register as well). > Of course it works on every platform. If you can have a thread-private > object, just store a pointer to an object living in a shared heap there. This assumes threads don't share all their memory. I thought your entire purpose in this thread was to keep things so that they do. > If one additional indirection is "a lot slower", then I don't know > what can be "a bit slower". I'm sorry, I thought we were trying to keep it so that all threads have the exact same view of memory. DS .