On Mon, Jun 16, 2014 at 2:20 PM, Stephane Viau sviau@codeaurora.org wrote:

This changes activates the iommu support for MDP5, through the platform config structure.

Iommu support is also slightly modified in order to make sure that MDP iommu is properly cleaned up if a probe deferral is requested. Before this change, IOMMU faults would occur if the probe failed (-EPROBE_DEFER).

So, this looks like it is really two patches.. one making -EPROBE_DEFER work properly, and the other adding the missing bits for mdp5 which so far are only on downstream kernel.

Could you split this patch into two, so I can queue the -EPROBE_DEFER fix for a 3.16 -fixes pull req?

BR, -R

Signed-off-by: Stephane Viau sviau@codeaurora.org

drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c | 28 +++++++++++++++++++++++----- drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h | 1 + drivers/gpu/drm/msm/msm_gem.c | 6 ++++++ drivers/gpu/drm/msm/msm_iommu.c | 21 +++++++++++++++++++-- drivers/gpu/drm/msm/msm_mmu.h | 1 + 5 files changed, 50 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c index 2967b19..62ee5cd 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c @@ -20,6 +20,10 @@ #include "msm_mmu.h" #include "mdp5_kms.h"

+static const char *iommu_ports[] = {

•           "mdp_0",
  

+};

static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev);

uint32_t __read_mostly _mdp5_pipe_vig_base; @@ -201,6 +205,12 @@ static void mdp5_preclose(struct msm_kms *kms, struct drm_file *file) static void mdp5_destroy(struct msm_kms *kms) { struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));

•   struct msm_mmu *mmu = mdp5_kms->mmu;
  

•   if (mmu) {
  

•           mmu->funcs->detach(mmu, iommu_ports, ARRAY_SIZE(iommu_ports));
  

•           mmu->funcs->destroy(mmu);
  

•   }
    kfree(mdp5_kms);
  

}

@@ -313,10 +323,6 @@ fail: return ret; }

-static const char *iommu_ports[] = {

•           "mdp_0",
  

-};

static int get_clk(struct platform_device *pdev, struct clk **clkp, const char *name) { @@ -406,17 +412,23 @@ struct msm_kms *mdp5_kms_init(struct drm_device *dev) mmu = msm_iommu_new(dev, config->iommu); if (IS_ERR(mmu)) { ret = PTR_ERR(mmu);

•                   dev_err(dev->dev, "failed to init iommu: %d\n", ret);
                    goto fail;
            }
  

•           ret = mmu->funcs->attach(mmu, iommu_ports,
                            ARRAY_SIZE(iommu_ports));
  

•           if (ret)
  

•           if (ret) {
  

•                   dev_err(dev->dev, "failed to attach iommu: %d\n", ret);
  

•                   mmu->funcs->destroy(mmu);
                    goto fail;
  

•           }
    } else {
            dev_info(dev->dev, "no iommu, fallback to phys "
                            "contig buffers for scanout\n");
            mmu = NULL;
    }
  

•   mdp5_kms->mmu = mmu;
  
    mdp5_kms->id = msm_register_mmu(dev, mmu);
    if (mdp5_kms->id < 0) {
  

@@ -445,5 +457,11 @@ static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev) #ifdef CONFIG_OF /* TODO */ #endif

•   config.iommu = iommu_domain_alloc(&platform_bus_type, 0);
  

•   /* TODO hard-coded in downstream mdss, but should it be? */
  

•   config.max_clk = 200000000;
  

•   /* TODO get from DT: */
  

•   config.smp_blk_cnt = 22;
  

•   return &config;
  

} diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h index 6a89b04..20ea748 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h @@ -41,6 +41,7 @@ struct mdp5_kms {

    /* mapper-id used to request GEM buffer mapped for scanout: */
    int id;

•   struct msm_mmu *mmu;
  
    /* for tracking smp allocation amongst pipes: */
    mdp5_smp_state_t smp_state;
  

diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index bb8026d..690d7e7 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -278,6 +278,7 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id, uint32_t *iova) { struct msm_gem_object *msm_obj = to_msm_bo(obj);

•   struct drm_device *dev = obj->dev;
    int ret = 0;
  
    if (!msm_obj->domain[id].iova) {
  

@@ -285,6 +286,11 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id, struct msm_mmu *mmu = priv->mmus[id]; struct page **pages = get_pages(obj);

•           if (!mmu) {
  

•                   dev_err(dev->dev, "null MMU pointer\n");
  

•                   return -EINVAL;
  

•           }
  

•           if (IS_ERR(pages))
                    return PTR_ERR(pages);
  
  

diff --git a/drivers/gpu/drm/msm/msm_iommu.c b/drivers/gpu/drm/msm/msm_iommu.c index 92b7459..198b2fe 100644 --- a/drivers/gpu/drm/msm/msm_iommu.c +++ b/drivers/gpu/drm/msm/msm_iommu.c @@ -28,7 +28,7 @@ static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev, unsigned long iova, int flags, void *arg) { DBG("*** fault: iova=%08lx, flags=%d", iova, flags);

•   return 0;
  

•   return -ENOSYS;
  

}

static int msm_iommu_attach(struct msm_mmu *mmu, const char **names, int cnt) @@ -40,8 +40,10 @@ static int msm_iommu_attach(struct msm_mmu *mmu, const char **names, int cnt) for (i = 0; i < cnt; i++) { struct device *msm_iommu_get_ctx(const char *ctx_name); struct device *ctx = msm_iommu_get_ctx(names[i]);

•           if (IS_ERR_OR_NULL(ctx))
  

•           if (IS_ERR_OR_NULL(ctx)) {
  

•                   dev_warn(dev->dev, "couldn't get %s context", names[i]);
                    continue;
  

•           }
            ret = iommu_attach_device(iommu->domain, ctx);
            if (ret) {
                    dev_warn(dev->dev, "could not attach iommu to %s", names[i]);
  

@@ -52,6 +54,20 @@ static int msm_iommu_attach(struct msm_mmu *mmu, const char **names, int cnt) return 0; }

+static void msm_iommu_detach(struct msm_mmu *mmu, const char **names, int cnt) +{

•   struct msm_iommu *iommu = to_msm_iommu(mmu);
  

•   int i;
  

•   for (i = 0; i < cnt; i++) {
  

•           struct device *msm_iommu_get_ctx(const char *ctx_name);
  

•           struct device *ctx = msm_iommu_get_ctx(names[i]);
  

•           if (IS_ERR_OR_NULL(ctx))
  

•                   continue;
  

•           iommu_detach_device(iommu->domain, ctx);
  

•   }
  

+}

static int msm_iommu_map(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt, unsigned len, int prot) { @@ -127,6 +143,7 @@ static void msm_iommu_destroy(struct msm_mmu *mmu)

static const struct msm_mmu_funcs funcs = { .attach = msm_iommu_attach,

•           .detach = msm_iommu_detach,
            .map = msm_iommu_map,
            .unmap = msm_iommu_unmap,
            .destroy = msm_iommu_destroy,
  

diff --git a/drivers/gpu/drm/msm/msm_mmu.h b/drivers/gpu/drm/msm/msm_mmu.h index 0303244..21da6d1 100644 --- a/drivers/gpu/drm/msm/msm_mmu.h +++ b/drivers/gpu/drm/msm/msm_mmu.h @@ -22,6 +22,7 @@

struct msm_mmu_funcs { int (*attach)(struct msm_mmu *mmu, const char **names, int cnt);

•   void (*detach)(struct msm_mmu *mmu, const char **names, int cnt);
    int (*map)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt,
                    unsigned len, int prot);
    int (*unmap)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt,
  

-- The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, hosted by The Linux Foundation