[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH 05/19] mm: when adding a region, merge with region after as w
From: |
Daniel Kiper |
Subject: |
Re: [PATCH 05/19] mm: when adding a region, merge with region after as well as before |
Date: |
Thu, 4 Nov 2021 17:36:52 +0100 |
User-agent: |
NeoMutt/20170113 (1.7.2) |
On Tue, Oct 12, 2021 at 06:29:54PM +1100, Daniel Axtens wrote:
> On x86_64-efi (at least) regions seem to be added from top down. The mm
> code will merge a new region with an existing region that comes
> immediately before the new region. This allows larger allocations to be
> satisfied that would otherwise be the case.
>
> On powerpc-ieee1275, however, regions are added from bottom up. So if
> we add 3x 32MB regions, we can still only satisfy a 32MB allocation,
> rather than the 96MB allocation we might otherwise be able to satisfy.
>
> * Define 'post_size' as being bytes lost to the end of an allocation
> due to being given weird sizes from firmware that are not multiples
> of GRUB_MM_ALIGN.
>
> * Allow merging of regions immediately _after_ existing regions, not
> just before. As with the other approach, we create an allocated
> block to represent the new space and the pass it to grub_free() to
> get the metadata right.
>
> Signed-off-by: Daniel Axtens <dja@axtens.net>
> ---
> grub-core/kern/mm.c | 55 +++++++++++++++++++++++++--------------
> include/grub/mm_private.h | 15 +++++++++++
> 2 files changed, 51 insertions(+), 19 deletions(-)
>
> diff --git a/grub-core/kern/mm.c b/grub-core/kern/mm.c
> index c070afc621f8..835ed8a8f6f9 100644
> --- a/grub-core/kern/mm.c
> +++ b/grub-core/kern/mm.c
> @@ -129,25 +129,41 @@ grub_mm_init_region (void *addr, grub_size_t size)
> size = ((grub_addr_t) -0x1000) - (grub_addr_t) addr;
>
> for (p = &grub_mm_base, q = *p; q; p = &(q->next), q = *p)
> - if ((grub_uint8_t *) addr + size + q->pre_size == (grub_uint8_t *) q)
> - {
> - r = (grub_mm_region_t) ALIGN_UP ((grub_addr_t) addr, GRUB_MM_ALIGN);
> - *r = *q;
> - r->pre_size += size;
> -
> - if (r->pre_size >> GRUB_MM_ALIGN_LOG2)
> - {
> - h = (grub_mm_header_t) (r + 1);
> - h->size = (r->pre_size >> GRUB_MM_ALIGN_LOG2);
> - h->magic = GRUB_MM_ALLOC_MAGIC;
> - r->size += h->size << GRUB_MM_ALIGN_LOG2;
> - r->pre_size &= (GRUB_MM_ALIGN - 1);
> - *p = r;
> - grub_free (h + 1);
> - }
> - *p = r;
> - return;
> - }
> + {
> + /* Does this region come _before_ an existing region? */
> + if ((grub_uint8_t *) addr + size + q->pre_size == (grub_uint8_t *) q)
> + {
> + r = (grub_mm_region_t) ALIGN_UP ((grub_addr_t) addr, GRUB_MM_ALIGN);
> + *r = *q;
> + r->pre_size += size;
> +
> + if (r->pre_size >> GRUB_MM_ALIGN_LOG2)
> + {
> + h = (grub_mm_header_t) (r + 1);
> + h->size = (r->pre_size >> GRUB_MM_ALIGN_LOG2);
> + h->magic = GRUB_MM_ALLOC_MAGIC;
> + r->size += h->size << GRUB_MM_ALIGN_LOG2;
> + r->pre_size &= (GRUB_MM_ALIGN - 1);
> + *p = r;
> + grub_free (h + 1);
> + }
> + *p = r;
> + return;
> + }
> +
> + /* Does this region come _after_ an existing region? */
> + if ((grub_uint8_t *)q + sizeof(*q) + q->size + q->post_size ==
> + (grub_uint8_t *) addr)
> + {
> + h = (grub_mm_header_t) ((grub_uint8_t *)addr - q->post_size);
> + h->size = (size + q->post_size) >> GRUB_MM_ALIGN_LOG2;
> + h->magic = GRUB_MM_ALLOC_MAGIC;
> + q->size += h->size << GRUB_MM_ALIGN_LOG2;
> + q->post_size = (q->post_size + size) & (GRUB_MM_ALIGN - 1);
> + grub_free (h + 1);
> + return;
The code itself LGTM. However, I would be more than happy if you add
bunch of comments and maybe some ASCII drawings explaining what is
happening here. This should save a lot time of our successors... :-)
> + }
> + }
>
> /* Allocate a region from the head. */
> r = (grub_mm_region_t) ALIGN_UP ((grub_addr_t) addr, GRUB_MM_ALIGN);
> @@ -166,6 +182,7 @@ grub_mm_init_region (void *addr, grub_size_t size)
> r->first = h;
> r->pre_size = (grub_addr_t) r - (grub_addr_t) addr;
> r->size = (h->size << GRUB_MM_ALIGN_LOG2);
> + r->post_size = size - r->size;
>
> /* Find where to insert this region. Put a smaller one before bigger ones,
> to prevent fragmentation. */
> diff --git a/include/grub/mm_private.h b/include/grub/mm_private.h
> index 533b47173e18..0effbc45a668 100644
> --- a/include/grub/mm_private.h
> +++ b/include/grub/mm_private.h
> @@ -74,8 +74,23 @@ typedef struct grub_mm_region
> */
> grub_size_t pre_size;
>
> + /* Likewise, the post-size is the number of bytes we wasted at the end
> + of the allocation because it wasn't a multiple of GRUB_MM_ALIGN
> + */
Please fix the formatting of this comment.
> + grub_size_t post_size;
> +
> /* How many bytes are in this region? (free and allocated) */
> grub_size_t size;
> +
> + /* pad to a multiple of cell size */
> +#if GRUB_CPU_SIZEOF_VOID_P == 4
> + char padding[4+4+4];
> +#elif GRUB_CPU_SIZEOF_VOID_P == 8
> + char padding[8+8+8];
> +#else
> +# error "unknown word size"
> +#endif
grub_uint8_t padding[3 * GRUB_CPU_SIZEOF_VOID_P];?
Or do not we have a construct in the C which allows us to add paddings
like that one in smarter way?
Daniel
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- Re: [PATCH 05/19] mm: when adding a region, merge with region after as well as before,
Daniel Kiper <=