Messages in this thread | | | Date | Fri, 27 Feb 2009 13:05:59 -0800 | From | "H. Peter Anvin" <> | Subject | Re: [PATCH RFC] x86: add brk allocation for very, very early allocations |
| |
Yinghai Lu wrote: > > it seems reserve _brk_end is some late? > > init_memory_mapping(0,...) could get some for direct mapping page table. and it could start from _end... >
That would make sense. init_memory_mapping() needs to know what not to step on, and that would include the brk at this stage. For 64-bit mode, Xen is the only user of !PSE, and so may be particularly vulnerable to this issue (because of massively larger direct mapping tables.)
This reminds me... is your intent that the BRK is permanent (unless explicitly freed on a page by page basis) or part of the init memory that is flushed?
-hpa
-- H. Peter Anvin, Intel Open Source Technology Center I work for Intel. I don't speak on their behalf.
| |