Re: [PATCH v4 3/4] x86/boot: Rewrite EFI/MBI2 code partly in C

[Frediano Ziglio <frediano.ziglio@xxxxxxxxx>]

On Thu, Sep 26, 2024 at 7:58 AM Jan Beulich <jbeulich@xxxxxxxx> wrote:
>
> On 25.09.2024 22:20, Andrew Cooper wrote:
> > On 25/09/2024 7:01 am, Frediano Ziglio wrote:
> >> +const char * asmlinkage __init
> >> +efi_multiboot2_prelude(uint32_t magic, const multiboot2_fixed_t *mbi)
> >> +{
> >> +    const multiboot2_tag_t *tag;
> >> +    EFI_HANDLE ImageHandle = NULL;
> >> +    EFI_SYSTEM_TABLE *SystemTable = NULL;
> >> +    const char *cmdline = NULL;
> >> +    bool have_bs = false;
> >> +
> >> +    if ( magic != MULTIBOOT2_BOOTLOADER_MAGIC )
> >> +        return "ERR: Not a Multiboot2 bootloader!";
> >> +
> >> +    /* Skip Multiboot2 information fixed part. */
> >> +    tag = _p(ROUNDUP((unsigned long)(mbi + 1), MULTIBOOT2_TAG_ALIGN));
> >> +
> >> +    for ( ; (const void *)tag - (const void *)mbi < mbi->total_size
> >> +            && tag->type != MULTIBOOT2_TAG_TYPE_END;
> >
> > && on previous line.
> >
> > But, this can move into the switch statement and reduce the for()
> > expression somewhat.
>
> While it can in principle, it would require further adjustments to the
> loop body, which I'm uncertain would be desirable. I notice Frediano
> also didn't make any change in v5 for this particular comment. Frediano:
> As indicated before, sending a new version without addressing all
> comments isn't nice. If you don't agree with a comment and hence don't
> make a change requested, you will want to "address" the comment verbally.
>
> Jan

I was going to reply, then I forgot. Yes, I'm trying to reply and
address all comments. In this specific instance, for some reason while
I was addressing the others, I forgot this.

Frediano