Message ID | 20221227-ux500-stm32-hash-v3-0-32ee12cd6f06@linaro.org (mailing list archive) |
---|---|
Headers | show |
Series | crypto: stm32 hash - reuse for Ux500 | expand |
On Wed, Jan 25, 2023 at 01:23:06AM +0100, Linus Walleij wrote: > By taking some small portions of the Ux500 HASH driver and > adding to the STM32 driver, it turns out we can support both > platforms with the more modern STM32 driver. > > The STM32 driver is more modern and compact thanks to using > things like the crypto engine. > > We add a polled mode since the Ux500 does not have any > interrupt. Incidentally, this could perhaps be re-used to > implement synchronous mode, if this is desireable. > > To: Herbert Xu <herbert@gondor.apana.org.au> > To: "David S. Miller" <davem@davemloft.net> > To: Rob Herring <robh+dt@kernel.org> > To: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> > To: Maxime Coquelin <mcoquelin.stm32@gmail.com> > To: Alexandre Torgue <alexandre.torgue@foss.st.com> > To: Lionel Debieve <lionel.debieve@foss.st.com> > Cc: linux-crypto@vger.kernel.org > Cc: devicetree@vger.kernel.org > Cc: linux-stm32@st-md-mailman.stormreply.com > Cc: linux-arm-kernel@lists.infradead.org > Cc: linux-kernel@vger.kernel.org > Signed-off-by: Linus Walleij <linus.walleij@linaro.org> > > --- > Changes in v3: > - Allocate the synchronous fallback algorithm in .cra_init() > and free it in .cra_exit(). > - Pick up some review tags. > - Link to v2: https://lore.kernel.org/r/20221227-ux500-stm32-hash-v2-0-bc443bc44ca4@linaro.org > > Changes in v2: > - Use an else-clause in the DT bindings. > - Fix up issues pointed out by Lionel in the driver extension. > - Dropped the patch converting dma_mode to a bool after > Lionel explained how this works. > - Link to v1: https://lore.kernel.org/r/20221227-ux500-stm32-hash-v1-0-b637ac4cda01@linaro.org > > --- > Linus Walleij (6): > dt-bindings: crypto: Let STM32 define Ux500 HASH > crypto: stm32/hash: Simplify code > crypto: stm32/hash: Use existing busy poll function > crypto: stm32/hash: Wait for idle before final CPU xmit > crypto: stm32/hash: Support Ux500 hash > crypto: ux500/hash - delete driver > > .../devicetree/bindings/crypto/st,stm32-hash.yaml | 23 +- > drivers/crypto/Kconfig | 10 - > drivers/crypto/Makefile | 1 - > drivers/crypto/stm32/stm32-hash.c | 266 ++- > drivers/crypto/ux500/Kconfig | 22 - > drivers/crypto/ux500/Makefile | 7 - > drivers/crypto/ux500/hash/Makefile | 11 - > drivers/crypto/ux500/hash/hash_alg.h | 398 ---- > drivers/crypto/ux500/hash/hash_core.c | 1966 -------------------- > 9 files changed, 249 insertions(+), 2455 deletions(-) > --- > base-commit: 1b929c02afd37871d5afb9d498426f83432e71c2 > change-id: 20221227-ux500-stm32-hash-9ee26834292f > > Best regards, > -- > Linus Walleij <linus.walleij@linaro.org> All applied. Thanks.
By taking some small portions of the Ux500 HASH driver and adding to the STM32 driver, it turns out we can support both platforms with the more modern STM32 driver. The STM32 driver is more modern and compact thanks to using things like the crypto engine. We add a polled mode since the Ux500 does not have any interrupt. Incidentally, this could perhaps be re-used to implement synchronous mode, if this is desireable. To: Herbert Xu <herbert@gondor.apana.org.au> To: "David S. Miller" <davem@davemloft.net> To: Rob Herring <robh+dt@kernel.org> To: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> To: Maxime Coquelin <mcoquelin.stm32@gmail.com> To: Alexandre Torgue <alexandre.torgue@foss.st.com> To: Lionel Debieve <lionel.debieve@foss.st.com> Cc: linux-crypto@vger.kernel.org Cc: devicetree@vger.kernel.org Cc: linux-stm32@st-md-mailman.stormreply.com Cc: linux-arm-kernel@lists.infradead.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Linus Walleij <linus.walleij@linaro.org> --- Changes in v3: - Allocate the synchronous fallback algorithm in .cra_init() and free it in .cra_exit(). - Pick up some review tags. - Link to v2: https://lore.kernel.org/r/20221227-ux500-stm32-hash-v2-0-bc443bc44ca4@linaro.org Changes in v2: - Use an else-clause in the DT bindings. - Fix up issues pointed out by Lionel in the driver extension. - Dropped the patch converting dma_mode to a bool after Lionel explained how this works. - Link to v1: https://lore.kernel.org/r/20221227-ux500-stm32-hash-v1-0-b637ac4cda01@linaro.org --- Linus Walleij (6): dt-bindings: crypto: Let STM32 define Ux500 HASH crypto: stm32/hash: Simplify code crypto: stm32/hash: Use existing busy poll function crypto: stm32/hash: Wait for idle before final CPU xmit crypto: stm32/hash: Support Ux500 hash crypto: ux500/hash - delete driver .../devicetree/bindings/crypto/st,stm32-hash.yaml | 23 +- drivers/crypto/Kconfig | 10 - drivers/crypto/Makefile | 1 - drivers/crypto/stm32/stm32-hash.c | 266 ++- drivers/crypto/ux500/Kconfig | 22 - drivers/crypto/ux500/Makefile | 7 - drivers/crypto/ux500/hash/Makefile | 11 - drivers/crypto/ux500/hash/hash_alg.h | 398 ---- drivers/crypto/ux500/hash/hash_core.c | 1966 -------------------- 9 files changed, 249 insertions(+), 2455 deletions(-) --- base-commit: 1b929c02afd37871d5afb9d498426f83432e71c2 change-id: 20221227-ux500-stm32-hash-9ee26834292f Best regards,