linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: matt@codeblueprint.co.uk (Matt Fleming)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v3 19/21] efi: stub: add implementation of efi_random_alloc()
Date: Thu, 21 Jan 2016 16:10:06 +0000	[thread overview]
Message-ID: <20160121161006.GF2510@codeblueprint.co.uk> (raw)
In-Reply-To: <1452518355-4606-21-git-send-email-ard.biesheuvel@linaro.org>

On Mon, 11 Jan, at 02:19:13PM, Ard Biesheuvel wrote:
> This implements efi_random_alloc(), which allocates a chunk of memory of
> a certain size at a certain alignment, and uses the random_seed argument
> it receives to randomize the offset of the allocation.

s/offset/address/ ?

I see what you're getting at with the word "offset" but ultimately,
this is a memory allocation function, and it returns an address.

"offset" implies to me that the implementation allocates a larger
memory chunk than is required and returns an address that is >= the
start of the bigger-than-required-allocation.

> This is implemented by iterating over the UEFI memory map, counting the
> number of suitable slots (aligned offsets) within each region, and picking
> a random number between 0 and 'number of slots - 1' to select the slot,
> This should guarantee that each possible offset is chosen equally likely.
> 
> Suggested-by: Kees Cook <keescook@chromium.org>
> Cc: Matt Fleming <matt@codeblueprint.co.uk>
> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
> ---
>  drivers/firmware/efi/libstub/efistub.h |  4 +
>  drivers/firmware/efi/libstub/random.c  | 85 ++++++++++++++++++++
>  2 files changed, 89 insertions(+)
> 
> diff --git a/drivers/firmware/efi/libstub/efistub.h b/drivers/firmware/efi/libstub/efistub.h
> index 206b7252b9d1..7a38e29da53d 100644
> --- a/drivers/firmware/efi/libstub/efistub.h
> +++ b/drivers/firmware/efi/libstub/efistub.h
> @@ -46,4 +46,8 @@ void efi_get_virtmap(efi_memory_desc_t *memory_map, unsigned long map_size,
>  efi_status_t efi_get_random_bytes(efi_system_table_t *sys_table,
>  				  unsigned long size, u8 *out);
>  
> +efi_status_t efi_random_alloc(efi_system_table_t *sys_table_arg,
> +			      unsigned long size, unsigned long align_bits,
> +			      unsigned long *addr, unsigned long random_seed);
> +
>  #endif
> diff --git a/drivers/firmware/efi/libstub/random.c b/drivers/firmware/efi/libstub/random.c
> index f539b1e31459..d4829824508c 100644
> --- a/drivers/firmware/efi/libstub/random.c
> +++ b/drivers/firmware/efi/libstub/random.c
> @@ -33,3 +33,88 @@ efi_status_t efi_get_random_bytes(efi_system_table_t *sys_table,
>  
>  	return rng->get_rng(rng, NULL, size, out);
>  }
> +
> +/*
> + * Return a weight for a memory entry depending on how many offsets it covers
> + * that are suitably aligned and supply enough room for the allocation.
> + */
> +static unsigned long get_entry_weight(efi_memory_desc_t *md, unsigned long size,
> +				      unsigned long align_bits)
> +{
> +	u64 start, end;
> +
> +	if (md->type != EFI_CONVENTIONAL_MEMORY)
> +		return 0;
> +
> +	if (!(md->attribute & EFI_MEMORY_WB))
> +		return 0;

This could do with a comment. When would EFI_CONVENTIONAL_MEMORY not
have this attribute capability in the memory map?

> +
> +	start = round_up(md->phys_addr, 1 << align_bits);
> +	end = round_down(md->phys_addr + md->num_pages * EFI_PAGE_SIZE - size,
> +			 1 << align_bits);
> +
> +	if (start >= end)
> +		return 0;
> +
> +	return (end - start) >> align_bits;
> +}
> +
> +/*
> + * The UEFI memory descriptors have a virtual address field that is only used
> + * when installing the virtual mapping using SetVirtualAddressMap(). Since it
> + * is unused here, we can reuse it to keep track of each descriptor's weight.
> + */
> +#define MD_WEIGHT(md)	((md)->virt_addr)
> +
> +efi_status_t efi_random_alloc(efi_system_table_t *sys_table_arg,
> +			      unsigned long size, unsigned long align_bits,
> +			      unsigned long *addr, unsigned long random_seed)
> +{
> +	unsigned long map_size, desc_size, max_weight = 0, target;
> +	efi_memory_desc_t *memory_map;
> +	efi_status_t status = EFI_NOT_FOUND;
> +	int l;

Could you pick a more descriptive variable name?

> +
> +	status = efi_get_memory_map(sys_table_arg, &memory_map, &map_size,
> +				    &desc_size, NULL, NULL);
> +	if (status != EFI_SUCCESS)
> +		return status;
> +
> +	/* assign each entry in the memory map a weight */
> +	for (l = 0; l < map_size; l += desc_size) {
> +		efi_memory_desc_t *md = (void *)memory_map + l;
> +		unsigned long weight;
> +
> +		weight = get_entry_weight(md, size, align_bits);
> +		MD_WEIGHT(md) = weight;
> +		max_weight += weight;
> +	}
> +
> +	/* find a random number between 0 and max_weight */
> +	target = (max_weight * (u16)random_seed) >> 16;
> +
> +	/* find the entry whose accumulated weight covers the target */
> +	for (l = 0; l < map_size; l += desc_size) {
> +		efi_memory_desc_t *md = (void *)memory_map + l;
> +
> +		if (target < MD_WEIGHT(md)) {
> +			unsigned long pages;
> +
> +			*addr = round_up(md->phys_addr, 1 << align_bits) +
> +				(target << align_bits);
> +			pages = round_up(size, EFI_PAGE_SIZE) / EFI_PAGE_SIZE;
> +
> +			status = efi_call_early(allocate_pages,
> +						EFI_ALLOCATE_ADDRESS,
> +						EFI_LOADER_DATA,
> +						pages,
> +						(efi_physical_addr_t *)addr);

You're mixing data types here. efi_physical_addr_t is always 64-bits,
but 'addr' is unsigned long, which is 32-bits on 32-bit platforms.
This cast isn't safe.

> +			break;
> +		}
> +		target -= MD_WEIGHT(md);

I think this needs a comment.

  reply	other threads:[~2016-01-21 16:10 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-01-11 13:18 [PATCH v3 00/21] arm64: implement support for KASLR Ard Biesheuvel
2016-01-11 13:18 ` [PATCH v3 01/21] of/fdt: make memblock minimum physical address arch configurable Ard Biesheuvel
2016-01-11 13:18 ` [PATCH v3 02/21] arm64: introduce KIMAGE_VADDR as the virtual base of the kernel region Ard Biesheuvel
2016-01-11 16:31   ` Mark Rutland
2016-01-11 13:18 ` [PATCH v3 03/21] arm64: pgtable: add dummy pud_index() and pmd_index() definitions Ard Biesheuvel
2016-01-11 17:40   ` Mark Rutland
2016-01-12 17:25     ` Ard Biesheuvel
2016-01-11 13:18 ` [PATCH v3 04/21] arm64: decouple early fixmap init from linear mapping Ard Biesheuvel
2016-01-11 16:09   ` Mark Rutland
2016-01-11 16:15     ` Ard Biesheuvel
2016-01-11 16:27       ` Mark Rutland
2016-01-11 16:51         ` Mark Rutland
2016-01-11 17:08           ` Ard Biesheuvel
2016-01-11 17:15             ` Ard Biesheuvel
2016-01-11 17:21               ` Mark Rutland
2016-01-11 13:18 ` [PATCH v3 05/21] arm64: kvm: deal with kernel symbols outside of " Ard Biesheuvel
2016-01-12 12:36   ` Mark Rutland
2016-01-12 13:23     ` Ard Biesheuvel
2016-01-11 13:18 ` [PATCH v3 06/21] arm64: pgtable: implement static [pte|pmd|pud]_offset variants Ard Biesheuvel
2016-01-11 16:24   ` Mark Rutland
2016-01-11 17:28     ` Ard Biesheuvel
2016-01-11 17:31       ` Mark Rutland
2016-01-11 13:19 ` [PATCH v3 07/21] arm64: move kernel image to base of vmalloc area Ard Biesheuvel
2016-01-12 18:14   ` Mark Rutland
2016-01-13  8:39     ` Ard Biesheuvel
2016-01-13  9:58       ` Ard Biesheuvel
2016-01-13 11:11         ` Mark Rutland
2016-01-13 11:14           ` Ard Biesheuvel
2016-01-13 13:51       ` Mark Rutland
2016-01-13 15:50         ` Ard Biesheuvel
2016-01-13 16:26           ` Mark Rutland
2016-01-14 18:57         ` Mark Rutland
2016-01-15  9:54           ` Ard Biesheuvel
2016-01-15 11:23             ` Mark Rutland
2016-01-27 14:31               ` Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 08/21] arm64: add support for module PLTs Ard Biesheuvel
2016-01-22 16:55   ` Mark Rutland
2016-01-22 17:06     ` Ard Biesheuvel
2016-01-22 17:19       ` Mark Rutland
2016-01-11 13:19 ` [PATCH v3 09/21] extable: add support for relative extables to search and sort routines Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 10/21] arm64: switch to relative exception tables Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 11/21] arm64: avoid R_AARCH64_ABS64 relocations for Image header fields Ard Biesheuvel
2016-01-13 18:12   ` Mark Rutland
2016-01-13 18:48     ` Ard Biesheuvel
2016-01-14  8:51       ` Ard Biesheuvel
2016-01-14  9:05         ` Ard Biesheuvel
2016-01-14 10:46           ` Mark Rutland
2016-01-14 11:22             ` Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 12/21] arm64: avoid dynamic relocations in early boot code Ard Biesheuvel
2016-01-14 17:09   ` Mark Rutland
2016-01-11 13:19 ` [PATCH v3 13/21] arm64: allow kernel Image to be loaded anywhere in physical memory Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 14/21] arm64: redefine SWAPPER_TABLE_SHIFT for use in asm code Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 14/21] arm64: [re]define SWAPPER_TABLE_[SHIFT|SIZE] " Ard Biesheuvel
2016-01-11 13:26   ` Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 15/21] arm64: split elf relocs into a separate header Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 16/21] scripts/sortextable: add support for ET_DYN binaries Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 17/21] arm64: add support for a relocatable kernel and KASLR Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 18/21] efi: stub: implement efi_get_random_bytes() based on EFI_RNG_PROTOCOL Ard Biesheuvel
2016-01-21 15:42   ` Matt Fleming
2016-01-21 16:12     ` Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 19/21] efi: stub: add implementation of efi_random_alloc() Ard Biesheuvel
2016-01-21 16:10   ` Matt Fleming [this message]
2016-01-21 16:16     ` Ard Biesheuvel
2016-01-11 13:19 ` [PATCH v3 20/21] efi: stub: use high allocation for converted command line Ard Biesheuvel
2016-01-21 16:20   ` Matt Fleming
2016-01-11 13:19 ` [PATCH v3 21/21] arm64: efi: invoke EFI_RNG_PROTOCOL to supply KASLR randomness Ard Biesheuvel
2016-01-21 16:31   ` Matt Fleming
2016-01-11 22:07 ` [PATCH v3 00/21] arm64: implement support for KASLR Kees Cook
2016-01-12  7:17   ` Ard Biesheuvel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160121161006.GF2510@codeblueprint.co.uk \
    --to=matt@codeblueprint.co.uk \
    --cc=linux-arm-kernel@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).