Re: [PATCH] mm/memblock: Add memblock_alloc_or_panic interface

[Weikang Guo <guoweikang.kernel@xxxxxxxxx>]

Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote on Saturday, 21
December 2024 07:06:
>
> On Fri, 20 Dec 2024 17:26:38 +0800 Guo Weikang <guoweikang.kernel@xxxxxxxxx> 
> wrote:
>
> > Before SLUB initialization, various subsystems used memblock_alloc to
> > allocate memory. In most cases, when memory allocation fails, an immediate
> > panic is required. To simplify this behavior and reduce repetitive checks,
> > introduce `memblock_alloc_or_panic`. This function ensures that memory
> > allocation failures result in a panic automatically, improving code
> > readability and consistency across subsystems that require this behavior.
> >
>
> Seems nice.
>
> > ...
> >
> > --- a/include/linux/memblock.h
> > +++ b/include/linux/memblock.h
> > @@ -417,6 +417,19 @@ static __always_inline void 
> > *memblock_alloc(phys_addr_t size, phys_addr_t align)
> >                                     MEMBLOCK_ALLOC_ACCESSIBLE, 
> > NUMA_NO_NODE);
> >  }
> >
> > +static __always_inline void *memblock_alloc_or_panic(phys_addr_t size, 
> > phys_addr_t align)
>
> We lost the printing of the function name, but it's easy to retain with
> something like
>
> #define memblock_alloc_or_panic(size, align)    \
>                 __memblock_alloc_or_panic(size, align, __func__)
>
You're absolutely right, this was an oversight on my part. I’ll make
sure to update it with the correct function name.
> > +{
> > +     void *addr = memblock_alloc(size, align);
> > +
> > +     if (unlikely(!addr))
> > +#ifdef CONFIG_PHYS_ADDR_T_64BIT
> > +             panic("%s: Failed to allocate %llu bytes\n", __func__, size);
>
> Won't this always print "memblock_alloc_or_panic: Failed ..."?  Not
> very useful.
>
As mentioned above.
> > +#else
> > +             panic("%s: Failed to allocate %u bytes\n", __func__, size);
> > +#endif
>
> We can avoid the ifdef with printk's "%pap"?
>
I appreciate you pointing this out. I wasn’t aware of this approach,
but it’s a great idea. It definitely simplifies things, and I’ve
learned something new in the process. I'll incorporate this into the
code.
> > +     return addr;
> > +}
>