Re: [PATCH container 3/3] setup: make the architecture fall back to amd64 for empty strings

[Fiona Ebner <f.ebner@proxmox.com>]

Am 04.02.26 um 10:17 AM schrieb Daniel Kral:
> Otherwise, if the underlying detect_architecture(...) method returns any
> false value, the return value of the call to protected_call(...) will

Do you mean undef value here? If I return 0 inside a protected call I get 0

> return an empty string.

not an empty string.

There seems to be a difference in behavior between being in a nested
protected call, which will return the result from the $sub directly, and
a non-nested protected call, which reads the result from the pipe, which
also results in an empty string when the result from $sub is undef.

I think the change here is fine, but we might want to document the
behavior for protected call. Or fix up the non-nested case to properly
return undef if $sub returns undef. Can be it's own series and will
require checking that use sites are happy with the change, but it seems
like there's only a small bunch that look at the return value anyways.

> This sets the architecture to an empty string and will make the
> container fail to start.
> 
> Signed-off-by: Daniel Kral <d.kral@proxmox.com>
> ---
>  src/PVE/LXC/Setup.pm | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/src/PVE/LXC/Setup.pm b/src/PVE/LXC/Setup.pm
> index 113093d..fb0207e 100644
> --- a/src/PVE/LXC/Setup.pm
> +++ b/src/PVE/LXC/Setup.pm
> @@ -153,7 +153,7 @@ sub new {
>              warn "Architecture detection failed: $err" if $err;
>          }
>  
> -        if (!defined($arch)) {
> +        if (!$arch) {
>              $arch = 'amd64';
>              print "Falling back to $arch.\nUse `pct set VMID --arch ARCH` to change.\n";
>          } else {