Re: [PATCH v2] [ifcombine] avoid dropping tree_could_trap_p [PR118514]

2025-01-23 Thread Richard Biener
On Fri, Jan 24, 2025 at 5:43 AM Alexandre Oliva wrote: > > On Jan 23, 2025, Richard Biener wrote: > > > That said, it'd be a lot clearer if this simply read > > > || !access_in_bounds_of_type_p (TREE_TYPE (inner), bs, bp) > > > without all the other weird conditions. > > ACK, will do. >

Re: [PATCH v2] [ifcombine] avoid dropping tree_could_trap_p [PR118514]

2025-01-23 Thread Alexandre Oliva
On Jan 23, 2025, Richard Biener wrote: > That said, it'd be a lot clearer if this simply read > || !access_in_bounds_of_type_p (TREE_TYPE (inner), bs, bp) > without all the other weird conditions. ACK, will do. >> + /* Check that the loads that we're trying to combine have the same

Re: [PATCH v2] [ifcombine] avoid dropping tree_could_trap_p [PR118514]

2025-01-23 Thread Richard Biener
On Thu, Jan 23, 2025 at 6:58 AM Alexandre Oliva wrote: > > On Jan 22, 2025, Alexandre Oliva wrote: > > > I have another patch coming up that doesn't raise concerns for me, so > > I'll hold off from installing the above, in case you also prefer the > > other one. > > Unlike other access patterns,

[PATCH v2] [ifcombine] avoid dropping tree_could_trap_p [PR118514]

2025-01-22 Thread Alexandre Oliva
On Jan 22, 2025, Alexandre Oliva wrote: > I have another patch coming up that doesn't raise concerns for me, so > I'll hold off from installing the above, in case you also prefer the > other one. Unlike other access patterns, BIT_FIELD_REFs aren't regarded as possibly-trapping out of referencing