From 51591baff67d787fd98ea134e55416ce4567804e Mon Sep 17 00:00:00 2001 From: Oliver Kowalke Date: Sun, 3 Feb 2013 19:45:53 +0100 Subject: [PATCH] update README.md --- README.md | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/ README.md b/ README.md index 12be349e..3da21790 100644 --- a/ README.md +++ b/ README.md @@ -4,7 +4,8 @@ boost.fiber boost.fiber provides a framework for micro-/userland-threads (fibers) scheduled cooperativly. The API contains classes and functions to manage and synchronize fibers similiar to boost.thread. This library is intended to support quasi-concurrency on embedded system or to replace boost.thread -for testing puposes (for instance checking for raise conditions). +for testing puposes (for instance checking for raise conditions) and to solve the many depended +task problem. A fiber is able to store the current execution state, including all registers and CPU flags, the instruction pointer, and the stack pointer and later restore this state. The idea is to have multiple @@ -12,6 +13,9 @@ execution paths running on a single thread using a sort of cooperative schedulin preemptively scheduled) - the running fiber decides explicitly when its yields to allow another fiber to run (context switching). +Fibers can be synchronized running in different threads. In order to support task steeling it boost.fiber +support migration of fibers between threads. + A context switch between threads costs usally thousends of CPU cycles on x86 compared to a fiber switch with less than 100 cycles. A fiber can only run on a single thread at any point in time but may be -migrated between threads. \ No newline at end of file +migrated between threads.