| Message ID | 1351046167-4882-6-git-send-email-mgreer@animalcreek.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Mark, "Mark A. Greer" <mgreer@animalcreek.com> writes: > From: "Mark A. Greer" <mgreer@animalcreek.com> > > Convert the omap-sham crypto driver to use the > pm_runtime API instead of the clk API. > > CC: Kevin Hilman <khilman@deeprootsystems.com> > CC: Paul Walmsley <paul@pwsan.com> > CC: Dmitry Kasatkin <dmitry.kasatkin@intel.com> > Signed-off-by: Mark A. Greer <mgreer@animalcreek.com> I can't pretend to fully understand this driver, It looks like the current code is doing a bit more fine-grained clock gating, and leaving the IP clocked only when needed. The proposed version does a 'get' in probe and a 'put' in remove, which means the IP will always be enabled (and thus preventing low-power states), even when it's not in use. If that's really needed, it should be thoroughly described in the changelog, otherwise I suggest doing the runtime PM 'get' and 'put' in roughtly the same spots as the current clk enable/disable which makes this a more straight-forward conversion. Kevin -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Oct 25, 2012 at 05:34:02PM -0700, Kevin Hilman wrote: > Hi Mark, Hi Kevin. > "Mark A. Greer" <mgreer@animalcreek.com> writes: > > > From: "Mark A. Greer" <mgreer@animalcreek.com> > > > > Convert the omap-sham crypto driver to use the > > pm_runtime API instead of the clk API. > > > > CC: Kevin Hilman <khilman@deeprootsystems.com> > > CC: Paul Walmsley <paul@pwsan.com> > > CC: Dmitry Kasatkin <dmitry.kasatkin@intel.com> > > Signed-off-by: Mark A. Greer <mgreer@animalcreek.com> > > I can't pretend to fully understand this driver, It looks like the > current code is doing a bit more fine-grained clock gating, and leaving > the IP clocked only when needed. > > The proposed version does a 'get' in probe and a 'put' in remove, which > means the IP will always be enabled (and thus preventing low-power > states), even when it's not in use. You're right, and what I did is wrong. I'll fix this in the next revision. Thanks, Mark -- -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/crypto/omap-sham.c b/drivers/crypto/omap-sham.c index a3fd6fc..acb85df 100644 --- a/drivers/crypto/omap-sham.c +++ b/drivers/crypto/omap-sham.c @@ -22,12 +22,12 @@ #include <linux/errno.h> #include <linux/interrupt.h> #include <linux/kernel.h> -#include <linux/clk.h> #include <linux/irq.h> #include <linux/io.h> #include <linux/platform_device.h> #include <linux/scatterlist.h> #include <linux/dma-mapping.h> +#include <linux/pm_runtime.h> #include <linux/delay.h> #include <linux/crypto.h> #include <linux/cryptohash.h> @@ -141,7 +141,6 @@ struct omap_sham_dev { struct device *dev; void __iomem *io_base; int irq; - struct clk *iclk; spinlock_t lock; int err; int dma; @@ -238,8 +237,6 @@ static void omap_sham_copy_ready_hash(struct ahash_request *req) static int omap_sham_hw_init(struct omap_sham_dev *dd) { - clk_enable(dd->iclk); - if (!test_bit(FLAGS_INIT, &dd->flags)) { omap_sham_write_mask(dd, SHA_REG_MASK, SHA_REG_MASK_SOFTRESET, SHA_REG_MASK_SOFTRESET); @@ -653,7 +650,6 @@ static void omap_sham_finish_req(struct ahash_request *req, int err) /* atomic operation is not needed here */ dd->flags &= ~(BIT(FLAGS_BUSY) | BIT(FLAGS_FINAL) | BIT(FLAGS_CPU) | BIT(FLAGS_DMA_READY) | BIT(FLAGS_OUTPUT_READY)); - clk_disable(dd->iclk); if (req->base.complete) req->base.complete(&req->base, err); @@ -1198,14 +1194,6 @@ static int __devinit omap_sham_probe(struct platform_device *pdev) if (err) goto dma_err; - /* Initializing the clock */ - dd->iclk = clk_get(dev, "ick"); - if (IS_ERR(dd->iclk)) { - dev_err(dev, "clock intialization failed.\n"); - err = PTR_ERR(dd->iclk); - goto clk_err; - } - dd->io_base = ioremap(dd->phys_base, SZ_4K); if (!dd->io_base) { dev_err(dev, "can't ioremap\n"); @@ -1213,11 +1201,12 @@ static int __devinit omap_sham_probe(struct platform_device *pdev) goto io_err; } - clk_enable(dd->iclk); + pm_runtime_enable(dev); + pm_runtime_get_sync(dev); + dev_info(dev, "hw accel on OMAP rev %u.%u\n", (omap_sham_read(dd, SHA_REG_REV) & SHA_REG_REV_MAJOR) >> 4, omap_sham_read(dd, SHA_REG_REV) & SHA_REG_REV_MINOR); - clk_disable(dd->iclk); spin_lock(&sham.lock); list_add_tail(&dd->list, &sham.dev_list); @@ -1235,9 +1224,8 @@ err_algs: for (j = 0; j < i; j++) crypto_unregister_ahash(&algs[j]); iounmap(dd->io_base); + pm_runtime_disable(dev); io_err: - clk_put(dd->iclk); -clk_err: omap_sham_dma_cleanup(dd); dma_err: if (dd->irq >= 0) @@ -1266,7 +1254,8 @@ static int __devexit omap_sham_remove(struct platform_device *pdev) crypto_unregister_ahash(&algs[i]); tasklet_kill(&dd->done_task); iounmap(dd->io_base); - clk_put(dd->iclk); + pm_runtime_put_sync(&pdev->dev); + pm_runtime_disable(&pdev->dev); omap_sham_dma_cleanup(dd); if (dd->irq >= 0) free_irq(dd->irq, dd);