[PATCH] mmc: store pointer to bio_crypt_ctx in mmc_request

[PATCH] mmc: store pointer to bio_crypt_ctx in mmc_request

[Eric Biggers]

From: Eric Biggers <ebiggers@google.com>

Make 'struct mmc_request' contain a pointer to the request's
'struct bio_crypt_ctx' directly, instead of extracting a 32-bit DUN from
it which is a cqhci-crypto specific detail.

This keeps the cqhci crypto specific details in the cqhci module, and it
makes mmc_core and mmc_block ready for MMC crypto hardware that accepts
the DUN and/or key in a way that is more flexible than that which will
be specified by the eMMC v5.2 standard.  Exynos SoCs are an example of
such hardware, as their inline encryption hardware takes keys directly
(it has no concept of keyslots) and supports 128-bit DUNs.

Note that the 32-bit DUN length specified by the standard is very
restrictive, so it is likely that more hardware will support longer DUNs
despite it not following the standard.  Thus, limiting the scope of the
32-bit DUN assumption to the place that actually needs it is warranted.

Signed-off-by: Eric Biggers <ebiggers@google.com>
---

This patch is already being carried in the Android Common Kernels, due
to it being needed by Exynos SoCs.  Unfortunately I have no way to
upstream Exynos eMMC inline encryption support at the moment, so this
patch isn't strictly required upstream at the moment.  But it may be
worthwhile as a cleanup regardless; it arguably fixes a layering
violation (mmc_core knowing about the details of cqhci crypto).

 drivers/mmc/core/crypto.c       | 15 ++++-----------
 drivers/mmc/host/cqhci-crypto.h |  7 +++++--
 include/linux/mmc/core.h        |  3 +--
 3 files changed, 10 insertions(+), 15 deletions(-)

diff --git a/drivers/mmc/core/crypto.c b/drivers/mmc/core/crypto.c
index 419a368f8402..67557808cada 100644
--- a/drivers/mmc/core/crypto.c
+++ b/drivers/mmc/core/crypto.c
@@ -31,18 +31,11 @@ void mmc_crypto_prepare_req(struct mmc_queue_req *mqrq)
 	struct request *req = mmc_queue_req_to_req(mqrq);
 	struct mmc_request *mrq = &mqrq->brq.mrq;
 
-	if (!req->crypt_keyslot)
+	if (!req->crypt_ctx)
 		return;
 
-	mrq->crypto_enabled = true;
-	mrq->crypto_key_slot = blk_ksm_get_slot_idx(req->crypt_keyslot);
-
-	/*
-	 * For now we assume that all MMC drivers set max_dun_bytes_supported=4,
-	 * which is the limit for CQHCI crypto.  So all DUNs should be 32-bit.
-	 */
-	WARN_ON_ONCE(req->crypt_ctx->bc_dun[0] > U32_MAX);
-
-	mrq->data_unit_num = req->crypt_ctx->bc_dun[0];
+	mrq->crypto_ctx = req->crypt_ctx;
+	if (req->crypt_keyslot)
+		mrq->crypto_key_slot = blk_ksm_get_slot_idx(req->crypt_keyslot);
 }
 EXPORT_SYMBOL_GPL(mmc_crypto_prepare_req);
diff --git a/drivers/mmc/host/cqhci-crypto.h b/drivers/mmc/host/cqhci-crypto.h
index 60b58ee0e625..d7fb084f563b 100644
--- a/drivers/mmc/host/cqhci-crypto.h
+++ b/drivers/mmc/host/cqhci-crypto.h
@@ -22,12 +22,15 @@ int cqhci_crypto_init(struct cqhci_host *host);
  */
 static inline u64 cqhci_crypto_prep_task_desc(struct mmc_request *mrq)
 {
-	if (!mrq->crypto_enabled)
+	if (!mrq->crypto_ctx)
 		return 0;
 
+	/* We set max_dun_bytes_supported=4, so all DUNs should be 32-bit. */
+	WARN_ON_ONCE(mrq->crypto_ctx->bc_dun[0] > U32_MAX);
+
 	return CQHCI_CRYPTO_ENABLE_BIT |
 	       CQHCI_CRYPTO_KEYSLOT(mrq->crypto_key_slot) |
-	       mrq->data_unit_num;
+	       mrq->crypto_ctx->bc_dun[0];
 }
 
 #else /* CONFIG_MMC_CRYPTO */
diff --git a/include/linux/mmc/core.h b/include/linux/mmc/core.h
index ab19245e9945..71101d1ec825 100644
--- a/include/linux/mmc/core.h
+++ b/include/linux/mmc/core.h
@@ -164,9 +164,8 @@ struct mmc_request {
 	int			tag;
 
 #ifdef CONFIG_MMC_CRYPTO
-	bool			crypto_enabled;
+	const struct bio_crypt_ctx *crypto_ctx;
 	int			crypto_key_slot;
-	u32			data_unit_num;
 #endif
 };
 
-- 
2.32.0.402.g57bb445576-goog

Re: [PATCH] mmc: store pointer to bio_crypt_ctx in mmc_request

[Ulf Hansson]

On Wed, 21 Jul 2021 at 17:48, Eric Biggers <ebiggers@kernel.org> wrote:
>
> From: Eric Biggers <ebiggers@google.com>
>
> Make 'struct mmc_request' contain a pointer to the request's
> 'struct bio_crypt_ctx' directly, instead of extracting a 32-bit DUN from
> it which is a cqhci-crypto specific detail.
>
> This keeps the cqhci crypto specific details in the cqhci module, and it
> makes mmc_core and mmc_block ready for MMC crypto hardware that accepts
> the DUN and/or key in a way that is more flexible than that which will
> be specified by the eMMC v5.2 standard.  Exynos SoCs are an example of
> such hardware, as their inline encryption hardware takes keys directly
> (it has no concept of keyslots) and supports 128-bit DUNs.
>
> Note that the 32-bit DUN length specified by the standard is very
> restrictive, so it is likely that more hardware will support longer DUNs
> despite it not following the standard.  Thus, limiting the scope of the
> 32-bit DUN assumption to the place that actually needs it is warranted.
>
> Signed-off-by: Eric Biggers <ebiggers@google.com>

Applied for next, thanks!

Kind regards
Uffe


> ---
>
> This patch is already being carried in the Android Common Kernels, due
> to it being needed by Exynos SoCs.  Unfortunately I have no way to
> upstream Exynos eMMC inline encryption support at the moment, so this
> patch isn't strictly required upstream at the moment.  But it may be
> worthwhile as a cleanup regardless; it arguably fixes a layering
> violation (mmc_core knowing about the details of cqhci crypto).
>
>  drivers/mmc/core/crypto.c       | 15 ++++-----------
>  drivers/mmc/host/cqhci-crypto.h |  7 +++++--
>  include/linux/mmc/core.h        |  3 +--
>  3 files changed, 10 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/mmc/core/crypto.c b/drivers/mmc/core/crypto.c
> index 419a368f8402..67557808cada 100644
> --- a/drivers/mmc/core/crypto.c
> +++ b/drivers/mmc/core/crypto.c
> @@ -31,18 +31,11 @@ void mmc_crypto_prepare_req(struct mmc_queue_req *mqrq)
>         struct request *req = mmc_queue_req_to_req(mqrq);
>         struct mmc_request *mrq = &mqrq->brq.mrq;
>
> -       if (!req->crypt_keyslot)
> +       if (!req->crypt_ctx)
>                 return;
>
> -       mrq->crypto_enabled = true;
> -       mrq->crypto_key_slot = blk_ksm_get_slot_idx(req->crypt_keyslot);
> -
> -       /*
> -        * For now we assume that all MMC drivers set max_dun_bytes_supported=4,
> -        * which is the limit for CQHCI crypto.  So all DUNs should be 32-bit.
> -        */
> -       WARN_ON_ONCE(req->crypt_ctx->bc_dun[0] > U32_MAX);
> -
> -       mrq->data_unit_num = req->crypt_ctx->bc_dun[0];
> +       mrq->crypto_ctx = req->crypt_ctx;
> +       if (req->crypt_keyslot)
> +               mrq->crypto_key_slot = blk_ksm_get_slot_idx(req->crypt_keyslot);
>  }
>  EXPORT_SYMBOL_GPL(mmc_crypto_prepare_req);
> diff --git a/drivers/mmc/host/cqhci-crypto.h b/drivers/mmc/host/cqhci-crypto.h
> index 60b58ee0e625..d7fb084f563b 100644
> --- a/drivers/mmc/host/cqhci-crypto.h
> +++ b/drivers/mmc/host/cqhci-crypto.h
> @@ -22,12 +22,15 @@ int cqhci_crypto_init(struct cqhci_host *host);
>   */
>  static inline u64 cqhci_crypto_prep_task_desc(struct mmc_request *mrq)
>  {
> -       if (!mrq->crypto_enabled)
> +       if (!mrq->crypto_ctx)
>                 return 0;
>
> +       /* We set max_dun_bytes_supported=4, so all DUNs should be 32-bit. */
> +       WARN_ON_ONCE(mrq->crypto_ctx->bc_dun[0] > U32_MAX);
> +
>         return CQHCI_CRYPTO_ENABLE_BIT |
>                CQHCI_CRYPTO_KEYSLOT(mrq->crypto_key_slot) |
> -              mrq->data_unit_num;
> +              mrq->crypto_ctx->bc_dun[0];
>  }
>
>  #else /* CONFIG_MMC_CRYPTO */
> diff --git a/include/linux/mmc/core.h b/include/linux/mmc/core.h
> index ab19245e9945..71101d1ec825 100644
> --- a/include/linux/mmc/core.h
> +++ b/include/linux/mmc/core.h
> @@ -164,9 +164,8 @@ struct mmc_request {
>         int                     tag;
>
>  #ifdef CONFIG_MMC_CRYPTO
> -       bool                    crypto_enabled;
> +       const struct bio_crypt_ctx *crypto_ctx;
>         int                     crypto_key_slot;
> -       u32                     data_unit_num;
>  #endif
>  };
>
> --
> 2.32.0.402.g57bb445576-goog
>