* pc-bios/optionrom: when/why do we build it?
@ 2024-11-22 6:58 Michael Tokarev
2024-11-22 7:05 ` Michael Tokarev
0 siblings, 1 reply; 3+ messages in thread
From: Michael Tokarev @ 2024-11-22 6:58 UTC (permalink / raw)
To: qemu-devel; +Cc: Paolo Bonzini
Hi!
Our configure includes pc-bios/optionrom/ into SUBDIRS variable in
config-host.mak when i386 compiler is available, so `make all` will
visit this subdir when building things. This will (re)build the
roms in pc-bios/optionroms/ directory, but wont update the binaries
in pc-bios/ which will be used to install things.
So the question is: why do we build them?
Thanks,
/mjt
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: pc-bios/optionrom: when/why do we build it?
2024-11-22 6:58 pc-bios/optionrom: when/why do we build it? Michael Tokarev
@ 2024-11-22 7:05 ` Michael Tokarev
2024-11-22 8:20 ` Paolo Bonzini
0 siblings, 1 reply; 3+ messages in thread
From: Michael Tokarev @ 2024-11-22 7:05 UTC (permalink / raw)
To: qemu-devel; +Cc: Paolo Bonzini
Maybe a more general question would be:
what is our main build entry point? Is it `make` (or `make all`)
still, or is it meson now?
For this specific optionrom thing, I'd say we should add a separate
rule to the top-level Makefile (or to meson.build) which is not
run by default but can be run explicitly, like `make optionrom` -
this will check if pc-bios/optionrom/config.mak is there and either
run make in this subdir or complain saying no i386 compiler is found.
Thanks,
/mjt
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: pc-bios/optionrom: when/why do we build it?
2024-11-22 7:05 ` Michael Tokarev
@ 2024-11-22 8:20 ` Paolo Bonzini
0 siblings, 0 replies; 3+ messages in thread
From: Paolo Bonzini @ 2024-11-22 8:20 UTC (permalink / raw)
To: Michael Tokarev, qemu-devel
On 11/22/24 08:05, Michael Tokarev wrote:
> Maybe a more general question would be:
> what is our main build entry point? Is it `make` (or `make all`)
> still, or is it meson now?
s/meson/ninja/
But no, it's always make. Using make ensures that:
1) you build stuff that is not emulators (option ROMs, TCG tests).
Meson is only for stuff that runs on the host
2) you rerun configure if it changes.
> For this specific optionrom thing, I'd say we should add a separate
> rule to the top-level Makefile (or to meson.build) which is not
> run by default but can be run explicitly, like `make optionrom` -
> this will check if pc-bios/optionrom/config.mak is there and either
> run make in this subdir or complain saying no i386 compiler is found.
The optionroms are built by default to ensure that they do not bitrot.
I think they're not copied (this is really old policy...) so that, if
there is no need, the binary does not change at all across releases.
But yes, adding an explicit makefile target is useful. It could also
copy the build products over to the source directory.
Paolo
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2024-11-22 8:21 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-11-22 6:58 pc-bios/optionrom: when/why do we build it? Michael Tokarev
2024-11-22 7:05 ` Michael Tokarev
2024-11-22 8:20 ` Paolo Bonzini
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).