On 25/02/2019 11:39, George Dunlap wrote: > On Fri, Feb 22, 2019 at 7:15 PM Andrew Cooper <andrew.coop...@citrix.com> > wrote: >> iremap_maddr and qinval_maddr point to the base of a block of contiguous RAM, >> allocated by the driver, holding the Interrupt Remapping table, and the >> Queued >> Invalidation ring. >> >> Despite their name, they are actually the values of the hardware register, >> including control metadata in the lower 12 bits. While uses of these fields >> do appear to correctly shift out the metadata, this is very subtle behaviour >> and confusing to follow. > Could I suggest that we add an ASSERT() to that macro, that maddr is > page aligned? > > The algorithm used certainly assumes that's the case, and would > produce incorrect results if it were ever validly true. Such an > ASSERT() would also have detected the storing-garbage-in-maddr issue > almost immediately.
Where would you suggest putting such an ASSERT() ? This variable is written exactly once. Irrespective, I plan to replace the existing scheme with a vmap()'d one which will delete this variable completely. ~Andrew _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
