[Toybox] toybox build first-world problems

[enh]

On Mon, Mar 6, 2023 at 8:15 PM Rob Landley <rob at landley.net> wrote:
>
> On 3/6/23 11:43, enh wrote:
> > (is there a reason this stuff isn't just `| xargs -P` though? macOS'
> > xargs does support -P, as does toybox's. might be worth a comment if
> > so.)
>
> I need to fetch the error return codes to see if any of the parallel build
> instances errored out, and the easy way to do that is for them to be shell jobs
> that I query the status of via wait.

ah, right. that makes sense. (at least if you ignore the fact that we
can't be the first people to want to run a bunch of jobs _and_ check
whether they succeeded or not!)

> Xargs -p doesn't stop if any of the commands return error, at least the one in
> debian discards stderr entirely, and it's also kinda weird about stdout too?
>
> $ echo -e 'one\ntwo\nthree\nfour\nfive' | xargs tee /dev/full
> $ ls one two three four five
> five  four  one  three  two
> $ echo -e 'one\ntwo\nthree\nfour\nfive' | xargs -P 3 bash -c 'echo $BASHPID; cat
> 2>&1'
> 28749
> $
>
> The first one created all six (individual) files (with no "/dev/full is full"
> complaints), but the second one only output one PID and I didn't dig into why
> (at a guess it closed stdout). I could presumably make a toybox version that did
> what I wanted, but then you'd need toybox to build toybox...
>
> > (and why does scripts/genconfig.sh take so much time if the .config
> > file hasn't changed? should genconfig() have a `-nt` test?)
>
> It has several, but some are commented out because the dependencies are...
> awkward. And I erred on the side of being reliable.
>
> You may remember last time I attacked that:
>
> http://lists.landley.net/pipermail/toybox-landley.net/2022-June/012906.html
>
> https://landley.net/notes-2022.html#:~:text=headers%20THREE%20TIMES
>
> The specific problem for .config has this comment:
>
> #TODO: "make $SED && make" doesn't regenerate config.h because diff .config
> if true #isnewer config.h "$KCONFIG_CONFIG"
>
> Which expands to "make sed" uses .singleconfig and redoes generated/* from that,
> and then if you do a normal toybox build next and it checks if generated/* is
> newer than ".config" the anwser is "yes" so it doesn't rebuild it, which is
> wrong. scripts/make.sh doesn't know if KCONFIG_CONFIG changed.

yes, incorrect is definitely a lot more annoying than slow (and that
one used to bite me a lot).

this weekend's changes made things 2x faster on the mac, so that
should stop me complaining for at least another year :-)

> It needs some design work (probably the same snapshot-the-variables trick cross
> compiler change detection is doing, just add KCONFIG_CONFIG). Hasn't made it
> back to the top of the todo pile yet...
>
> The two hard problems in computer science: naming things, cache invalidation,
> and fencepost errors.
>
> Rob