Re: [PATCH] entry: check for fstat() errors after checkout

[Derrick Stolee <stolee@gmail.com>]

On 7/21/2020 4:00 PM, Junio C Hamano wrote:
> Matheus Tavares Bernardino <matheus.bernardino@usp.br> writes:
> 
>> I was looking further at this code and noticed that the conditions
>> under which we fstat() (or lstat()) an entry are slightly different
>> throughout entry.c:
>>
>> - In write_entry()'s footer, we call lstat() iff stat->refresh_cache.
>> - In write_entry()'s `write_file_entry` label, we call fstat_output()
>> when !to_tempfile.
>> - In streaming_write_entry() we call fstat_output() without checking
>> if !to_tempfile.
>> - And, finally, in fstat_output() itself, we check
>> `state->refresh_cache && !state->base_dir_len`.
>>
>> I understand we always check state->refresh_cache to avoid getting
>> stat information we won't really need later, as we are not updating
>> the index. But why do we check !to_tempfile and !state->base_dir_len?
>> Doesn't writing to a tempfile or using a checkout prefix already imply
>> !state->refresh_cache?
> 
> You can easily blame the code back to e4c72923 (write_entry(): use
> fstat() instead of lstat() when file is open, 2009-02-09).  Back
> then, only a single place assigned 0 to state.refresh_cache and that
> is in "checkout-index" with either base_dir_len or to_tempfile set.
> 
> I do not remember, and I am fairly sure Stolee does not remember
> either.  If I have to guess, this was done merely to be extra
> cautious, perhaps? As refresh_cache bit is checked first, check for
> !to_tempfile and !base_dir_len would be dead at best and redundant
> at worst.

Yeah, this portion is way outside of my expertise. I'm happy
to _try_ reading patches, but I'd have difficulty being
confident in any change in this area.

Thanks,
-Stolee