Hi, I have some questions about uClinux page cache and memory reclamation. Probably some one can give some advice here. Thanks in advance.
My platform: ARM dev board + uclinux 2.6.25 My questions: If we limit the size of page caches or increase the write-back frequency for dirty data in page caches, can we avoid/decrease the frequency of OOM-Killer occurrence in the case of kernel memory being in pressure? I know the page cache is tunable, but is it helpful if I just want to avoid OOM-Killer happens. The implementation of page allocation function in kernel (i.e. __alloc_pages() in mm/page_alloc.c ) has called wakeup_kswapd() when get_page_from_freelist() function can't return a valid page, I don't know whether this function starts pdflush thread to reclaim memory? If so, does it indicate that it's not necessary for us to tune page caches? And is the tuning operation on page cache only for write-heavy loads, not for memory saving? -- Best Regards Erjian
_______________________________________________ uClinux-dev mailing list uClinux-dev@uclinux.org http://mailman.uclinux.org/mailman/listinfo/uclinux-dev This message was resent by uclinux-dev@uclinux.org To unsubscribe see: http://mailman.uclinux.org/mailman/options/uclinux-dev
