Tim Allen 0ab0ab6957 Merge commit '5e109224fb3480d806e44d74587d052c7f88bf83' into master 4 rokov pred
..
doc d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
LICENSE d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
README.md d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
aarch64.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
amd64.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
arm.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
fiber.c 0ab0ab6957 Merge commit '5e109224fb3480d806e44d74587d052c7f88bf83' into master 4 rokov pred
libco.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
libco.h d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
ppc.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
ppc64v2.c 0ab0ab6957 Merge commit '5e109224fb3480d806e44d74587d052c7f88bf83' into master 4 rokov pred
settings.h 0ab0ab6957 Merge commit '5e109224fb3480d806e44d74587d052c7f88bf83' into master 4 rokov pred
sjlj.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
ucontext.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred
x86.c d72874fa59 Merge libco repository's commit 'd31c6e75a54be12307d65fd80f55adfbe7aa9d67' into master 4 rokov pred

README.md

libco

libco is a cooperative multithreading library written in C89.

Although cooperative multithreading is limited to a single CPU core, it scales substantially better than preemptive multithreading.

For applications that need 100,000 or more context switches per second, the kernel overhead involved in preemptive multithreading can end up becoming the bottleneck in the application. libco can easily scale to 10,000,000 or more context switches per second.

Ideal use cases include servers (HTTP, RDBMS) and emulators (CPU cores, etc.)

It currently includes backends for:

  • x86 CPUs
  • amd64 CPUs
  • PowerPC CPUs
  • PowerPC64 ELFv1 CPUs
  • PowerPC64 ELFv2 CPUs
  • ARM 32-bit CPUs
  • ARM 64-bit (AArch64) CPUs
  • POSIX platforms (setjmp)
  • Windows platforms (fibers)

See [doc/targets.md] for details.

See [doc/usage.md] for documentation.

License

libco is released under the ISC license.