Digest for comp.programming.threads@googlegroups.com - 2 updates in 2 topics

comp.programming.threads@googlegroups.com

Google Groups

Topic digest
View all topics

• About my scalable MLock and scalable AMLock.. - 1 Update
• x86 Reordering - 1 Update

About my scalable MLock and scalable AMLock..

Horizon68 <horizon@horizon.com>: Jul 15 02:35PM -0700 Hello..     About my scalable MLock and scalable AMLock..     As i said:   x86 Reordering   All modern x86 chips from Intel and AMD aggressively reorder instructions across a window which is around 200 instructions deep on recent CPUs from both manufacturers (i.e. a new instruction may execute while an older instruction more than 200 instructions "in the past" is still waiting). This is generally all invisible to a single thread since the CPU still maintains the illusion of serial execution by the current thread by respecting dependencies, so from the point of view of the current thread of execution it is as-if the instructions were executed serially.     Here is my scalable MLock here:   https://sites.google.com/site/scalable68/scalable-mlock     And here is my scalable AMLock here:   https://sites.google.com/site/scalable68/scalable-amlock       I think i will soon enhance more my two scalable lock above and i will easily port them to ARM architecture, and now they are working with x86 architecture, and after that, my SemaMonitor and my SemaCondvar and my scalable RWLocks will be ported easily to ARM, and i will soon implement a portable Semaphore that uses my SemaCondvar and uses my portable Mutex that uses my SemaMonitor, they will soon all work with x86 and ARM.     Thank you, Amine Moulay Ramdane.

Back to top

x86 Reordering

Horizon68 <horizon@horizon.com>: Jul 15 10:42AM -0700 Hello..     x86 Reordering   All modern x86 chips from Intel and AMD aggressively reorder instructions across a window which is around 200 instructions deep on recent CPUs from both manufacturers (i.e. a new instruction may execute while an older instruction more than 200 instructions "in the past" is still waiting). This is generally all invisible to a single thread since the CPU still maintains the illusion of serial execution by the current thread by respecting dependencies, so from the point of view of the current thread of execution it is as-if the instructions were executed serially.     So i think the memory barriers positions of my scalable MLock (that is a scalable lock) are ok, read about it here:     https://sites.google.com/site/scalable68/scalable-mlock     Thank you, Amine Moulay Ramdane.

Back to top

You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to comp.programming.threads+unsubscribe@googlegroups.com.