![]() ![]() > The interrupt stacks are stackable, so we could use multiple > interrupt handlers that other kernel code will not eat into. > system when under heavy irq load by provide a fixed stack size for > with large numbers of tasks, and increases the reliability of the > on the VM subsystem, reduces the overall memory usage for systems > are a couple of reasons for doing this: 4KB stacks put less pressure > tasks, plus a seperate 4KB irq stack for use by interrupts. > Below is a patch against 2.5.20 that implements 4KB stacks for Need to disable preemption in common_interrupt() like it was before. ![]() The patch is good enough to pass LTP, butĬonsistently freezes when I run tcpdump on it.Īlthough I don't have CONFIG_PREEMPT on, I have the feeling that I I've resynced Ben's patch against 2.5.40. Reply: Andreas Dilger: "Re: 4KB stack + irq stack for x86". ![]() Reply: Benjamin LaHaise: "Re: 4KB stack + irq stack for x86".Next in thread: Benjamin LaHaise: "Re: 4KB stack + irq stack for x86".Previous message: Jeff Dike: " UML bug fixes".Next message: Petr Vandrovec: "Re: include/asm/irq_vectors.h not found".Linux-Kernel Archive: 4KB stack + irq stack for x 4KB stack + irq stack for x86 From: Dave Hansen ( Wed 16:18:36 EST
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |