Hi Christoph, I'm so sorry for the too late response as I took a longer vacations. On 01/31 2014 23:44 PM, Christoph Lameter wrote: > On Wed, 15 Jan 2014, Jeff Liu wrote: > >> Now I have a rough/stupid idea to add an extension to the slab caches [2], that is >> if the slab cache size is limited which could be determined in cache_grow(), the >> shrinker would be triggered accordingly. I'd like to learn/know if there are any >> suggestions and similar requirements in other subsystems. > > Hmmm.... Looks like you got the right point where to insert the code to > check for the limit. But lets leave the cache creation API the way it is. > Add a function to set the limit? Good idea. Yeah, changing the existing API is suboptimal than adding a new one. In this case, another thing I'm hesitating about whether to export the cache_limit via /proc/slabinfo by extending its tunable fields -- the per-CPU slab cache limit and batchcount, as thus will change the user space interface and slabtop(1) need to be modified accordingly. Thank! -Jeff -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>