On 18/04/16 14:44, Alexander Monakov wrote: > On Mon, 18 Apr 2016, Ramana Radhakrishnan wrote: >> On Mon, Apr 18, 2016 at 2:26 PM, Alexander Monakov <amona...@ispras.ru> >> wrote: >>> On Thu, 14 Apr 2016, Szabolcs Nagy wrote: >>>> looking at [2] i don't see why >>>> >>>> func: >>>> mov x9, x30 >>>> bl _tracefunc >>>> <function body> >>>> >>>> is not good for the kernel. >>>> >>>> mov x9, x30 is a nop at function entry, so in >>>> theory 4 byte atomic write should be enough >>>> to enable/disable tracing. >>> >>> Overwriting x9 can be problematic because GCC has gained the ability to >>> track >>> register usage interprocedurally: if foo() calls bar(), and GCC has already >>> emitted code for bar() and knows that it cannot change x9, it can use that >>> knowledge to avoid saving/restoring x9 in foo() around calls to bar(). See >>> option '-fipa-ra'. >>> >>> If there's no register that can be safely used in place of x9 here, then >>> the backend should emit the entry/pad appropriately (e.g. with an unspec >>> that >>> clobbers the possibly-overwritten register). >> >> Can you not use one of x16 / x17 the intra-procedure-call scratch >> registers as per the PCS ? > > As long as: > > - there's a guarantee that the call to _tracefunc wouldn't need some kind of > veneer that would clobber those; I don't know enough about AArch64 to say; >
A veneer could clobber them, yeah . > - and GCC is not smart enough to be aware that intra-TU calls to 'func' (the > function we're instrumenting) don't touch x16/x17. And GCC should be that > smart, if it's not, it's a bug, right? :) > That it already is - IIRC. Otherwise -fipa-ra wouldn't work . Alternatively just add x9 to the list for fipa-ra the same way as x16 / x17 are handled already , no ? Ramana > Thanks. > Alexander >
