> Am 13.06.2025 um 19:14 schrieb Paul Eggert <egg...@cs.ucla.edu>:
>
> On 2025-06-13 01:03, Peter Dyballa wrote:
>> Another question seems to be, why is the null pointer used when calling
>> fwrite()? A check should avoid this, rather early, and report or set some
>> failure status…
>
> No, that would be a bad way to "fix" things. We should find the real problem
> and fix it, not paper over it.
Correct, the actual cause for an error condition that was not thought of should
be determined.
But, when this case is found and fixed, one would again *assume* that no null
pointer will be passed to fwrite(). Until a new case appears.
There is so much code that checks for particular circumstances (could be
repeatedly – isn't there a large struct for every file of a set to diff/comp it
that could provide "regular file", "unreadable", "binary", "directory", etc.
condition from first check on?), so there could be as well one more, final,
check. And then write to stderr? The differences, or equal text blocks, seem to
be determined correctly, so why throw that data away? This is not economical.
I think debugging will continue a few days, probably. A bit confusing is that
functions are called from a handful of different files and Gdb is jumping
around… At least I think that when coming close to the failure I'd have the
series of breakpoints decimated, not stopping at each function call and
stepping into it!
--
Greetings
Pete
If it dies, it's biology. If it blows up, it's chemistry. If it doesn't work,
it's physics.
– University washroom sgraffito
