On 09/27/2012 01:34 PM, Aurelien Jarno wrote: >> (1) Helper may read globals. Sometimes indirectly via exception. >> Implies that globals must be synced, but may remain in REG/CONST state. > > I actually do wonder if we shouldn't use a different flag for "may > trigger exception", even if we end-up handling it the same way as "may > read globals". I am not sure it would be clear without knowledge of the > QEMU internals that a function triggering exceptions should be flagged > as "read globals".
I agree it is better to have a "may trigger exception" flag if for nothing else besides documentation. > Given the default is READG | WRITEG, it would be difficult to do that > unless we negate the flags. Actually NOWRITEG, NOREADG, NOSIDE might be > a good way to go, and is more consistent that by default nothing should > be implied. Sure. r~
