Li, Xin B
2006-Feb-27 12:11 UTC
[Xen-devel] Will map_domain_page return NULL when fails on x86_32?
Hi Keir, just a curious question, will map_domain_page return NULL when fails on x86_32? If not, why? thanks -Xin _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2006-Feb-27 13:17 UTC
[Xen-devel] Re: Will map_domain_page return NULL when fails on x86_32?
On 27 Feb 2006, at 12:11, Li, Xin B wrote:> Hi Keir, just a curious question, will map_domain_page return NULL when > fails on x86_32? If not, why? > thanksI don''t expect that it should ever fail. It''s used for temporary mappings (e.g., scope of a function) and so even though the mapping space is finite, we shouldn''t ever exhaust it. The map_domain_page_global() space is a different matter... -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Li, Xin B
2006-Feb-27 13:34 UTC
[Xen-devel] RE: Will map_domain_page return NULL when fails on x86_32?
>> Hi Keir, just a curious question, will map_domain_page >return NULL when >> fails on x86_32? If not, why? >> thanks > >I don''t expect that it should ever fail. It''s used for temporary >mappings (e.g., scope of a function) and so even though the mapping >space is finite, we shouldn''t ever exhaust it.So the unmap is not a must? Can you explain the mechanism of garbage collection here? Thanks -Xin> >The map_domain_page_global() space is a different matter... > > -- Keir >_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2006-Feb-27 13:48 UTC
[Xen-devel] Re: Will map_domain_page return NULL when fails on x86_32?
On 27 Feb 2006, at 13:34, Li, Xin B wrote:>>> Hi Keir, just a curious question, will map_domain_page >> return NULL when >>> fails on x86_32? If not, why? >>> thanks >> >> I don''t expect that it should ever fail. It''s used for temporary >> mappings (e.g., scope of a function) and so even though the mapping >> space is finite, we shouldn''t ever exhaust it. > > So the unmap is not a must? > Can you explain the mechanism of garbage collection here? > ThanksThe unmap is a must. After unmap the mapping is put in a per-vcpu lookup table to allow fast reuse. When it is evicted from that table then it is freed. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel