Re: [PATCH] c++: ICE with RANGE_EXPR and array init [PR109431]

2025-03-01 Thread Jason Merrill
On 2/28/25 12:16 PM, Marek Polacek wrote: Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk/branches? Yes. -- >8 -- We crash because we generate {[0 ... 1]={.low=0, .high=1}, [1]={.low=0, .high=1}} which output_constructor_regular_field doesn't want to see. This happens since

[PATCH] c++: ICE with RANGE_EXPR and array init [PR109431]

2025-02-28 Thread Marek Polacek
Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk/branches? -- >8 -- We crash because we generate {[0 ... 1]={.low=0, .high=1}, [1]={.low=0, .high=1}} which output_constructor_regular_field doesn't want to see. This happens since r9-1483: process_init_constructor_array can now creat