On Mon, Oct 21, 2019 at 01:15:21PM +0200, Christian König wrote:
> This patch is a stripped down version of the locking changes
> necessary to support dynamic DMA-buf handling.
>
> It adds a dynamic flag for both importers as well as exporters
> so that drivers can choose if they want the reservation object
> locked or unlocked during mapping of attachments.
>
> For compatibility between drivers we cache the DMA-buf mapping
> during attaching an importer as soon as exporter/importer
> disagree on the dynamic handling.
>
> This change has gone through a lengthy discussion on dri-devel
> and other mailing lists with at least 3-4 different attempts and
> dead-ends until we settled on this solution. Please refer to the
> mailing lists archives for full background on the history of
> this change.
I kinda hoped for a real write-up of why we ended up here, not a "please
read the last year or so of dri-devel" ...
So here's what I think we need to minimally mention, pls add:
<cut>
Issues and solutions we considered:
- We can't change all existing drivers, and existing drivers have strong
opinions about which locks they're holding while calling
dma_buf_attachment_map/unmap. The solution to avoid this was to move the
actual map/unmap out from this call, into the attach/detach callbacks,
and cache the mapping. This works because drivers don't call
attach/detach from deep within their code callchains (like deep in
memory management code called from cs/execbuf ioctl), but directly from
the fd2handle implementation.
- The caching has some troubles on some soc drivers, which set other modes
than DMA_BIDIRECTIONAL. We can't have 2 incompatible mappings, and we
can't re-create the mapping at _map time due to the above locking fun.
We very carefuly step around that by only caching at attach time if the
dynamic mode between importer/expoert mismatches.
- There's been quite some discussion on dma-buf mappings which need active
cache management, which would all break down when caching, plus we don't
have explicit flush operations on the attachment side. The solution to
this was to shrug and keep the current discrepancy between what the
dma-buf docs claim and what implementations do, with the hope that the
begin/end_cpu_access hooks are good enough and that all necessary
flushing to keep device mappings consistent will be done there.
</cut>
>
> v2: cleanup set_name merge, improve kerneldoc
>
> Signed-off-by: Christian König <christian.koenig@xxxxxxx>
> ---
> drivers/dma-buf/dma-buf.c | 102 +++++++++++++++++++++++++++++++++-----
> include/linux/dma-buf.h | 57 +++++++++++++++++++--
> 2 files changed, 143 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
> index 433d91d710e4..753be84b5fd6 100644
> --- a/drivers/dma-buf/dma-buf.c
> +++ b/drivers/dma-buf/dma-buf.c
> @@ -45,10 +45,10 @@ static char *dmabuffs_dname(struct dentry *dentry, char *buffer, int buflen)
> size_t ret = 0;
>
> dmabuf = dentry->d_fsdata;
> - mutex_lock(&dmabuf->lock);
> + dma_resv_lock(dmabuf->resv, NULL);
> if (dmabuf->name)
> ret = strlcpy(name, dmabuf->name, DMA_BUF_NAME_LEN);
> - mutex_unlock(&dmabuf->lock);
> + dma_resv_unlock(dmabuf->resv);
>
> return dynamic_dname(dentry, buffer, buflen, "/%s:%s",
> dentry->d_name.name, ret > 0 ? name : "");
> @@ -334,7 +334,7 @@ static long dma_buf_set_name(struct dma_buf *dmabuf, const char __user *buf)
> if (IS_ERR(name))
> return PTR_ERR(name);
>
> - mutex_lock(&dmabuf->lock);
> + dma_resv_lock(dmabuf->resv, NULL);
> if (!list_empty(&dmabuf->attachments)) {
> ret = -EBUSY;
> kfree(name);
> @@ -344,7 +344,7 @@ static long dma_buf_set_name(struct dma_buf *dmabuf, const char __user *buf)
> dmabuf->name = name;
>
> out_unlock:
> - mutex_unlock(&dmabuf->lock);
> + dma_resv_unlock(dmabuf->resv);
> return ret;
> }
>
> @@ -403,10 +403,10 @@ static void dma_buf_show_fdinfo(struct seq_file *m, struct file *file)
> /* Don't count the temporary reference taken inside procfs seq_show */
> seq_printf(m, "count:\t%ld\n", file_count(dmabuf->file) - 1);
> seq_printf(m, "exp_name:\t%s\n", dmabuf->exp_name);
> - mutex_lock(&dmabuf->lock);
> + dma_resv_lock(dmabuf->resv, NULL);
> if (dmabuf->name)
> seq_printf(m, "name:\t%s\n", dmabuf->name);
> - mutex_unlock(&dmabuf->lock);
> + dma_resv_unlock(dmabuf->resv);
> }
>
> static const struct file_operations dma_buf_fops = {
> @@ -525,6 +525,10 @@ struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info)
> return ERR_PTR(-EINVAL);
> }
>
> + if (WARN_ON(exp_info->ops->cache_sgt_mapping &&
> + exp_info->ops->dynamic_mapping))
> + return ERR_PTR(-EINVAL);
> +
> if (!try_module_get(exp_info->owner))
> return ERR_PTR(-ENOENT);
>
> @@ -645,10 +649,11 @@ void dma_buf_put(struct dma_buf *dmabuf)
> EXPORT_SYMBOL_GPL(dma_buf_put);
>
> /**
> - * dma_buf_attach - Add the device to dma_buf's attachments list; optionally,
> + * dma_buf_dynamic_attach - Add the device to dma_buf's attachments list; optionally,
> * calls attach() of dma_buf_ops to allow device-specific attach functionality
> - * @dmabuf: [in] buffer to attach device to.
> - * @dev: [in] device to be attached.
> + * @dmabuf: [in] buffer to attach device to.
> + * @dev: [in] device to be attached.
> + * @dynamic_mapping: [in] calling convention for map/unmap
> *
> * Returns struct dma_buf_attachment pointer for this attachment. Attachments
> * must be cleaned up by calling dma_buf_detach().
> @@ -662,8 +667,9 @@ EXPORT_SYMBOL_GPL(dma_buf_put);
> * accessible to @dev, and cannot be moved to a more suitable place. This is
> * indicated with the error code -EBUSY.
> */
> -struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf,
> - struct device *dev)
> +struct dma_buf_attachment *
> +dma_buf_dynamic_attach(struct dma_buf *dmabuf, struct device *dev,
> + bool dynamic_mapping)
> {
> struct dma_buf_attachment *attach;
> int ret;
> @@ -677,6 +683,7 @@ struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf,
>
> attach->dev = dev;
> attach->dmabuf = dmabuf;
> + attach->dynamic_mapping = dynamic_mapping;
>
> mutex_lock(&dmabuf->lock);
>
> @@ -685,16 +692,64 @@ struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf,
> if (ret)
> goto err_attach;
> }
> + dma_resv_lock(dmabuf->resv, NULL);
> list_add(&attach->node, &dmabuf->attachments);
> + dma_resv_unlock(dmabuf->resv);
>
> mutex_unlock(&dmabuf->lock);
>
> + /* When either the importer or the exporter can't handle dynamic
> + * mappings we cache the mapping here to avoid issues with the
> + * reservation object lock.
> + */
> + if (dma_buf_attachment_is_dynamic(attach) !=
> + dma_buf_is_dynamic(dmabuf)) {
> + struct sg_table *sgt;
> +
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_lock(attach->dmabuf->resv, NULL);
> +
> + sgt = dmabuf->ops->map_dma_buf(attach, DMA_BIDIRECTIONAL);
> + if (!sgt)
> + sgt = ERR_PTR(-ENOMEM);
> + if (IS_ERR(sgt)) {
> + ret = PTR_ERR(sgt);
> + goto err_unlock;
> + }
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_unlock(attach->dmabuf->resv);
> + attach->sgt = sgt;
> + attach->dir = DMA_BIDIRECTIONAL;
> + }
> +
> return attach;
>
> err_attach:
> kfree(attach);
> mutex_unlock(&dmabuf->lock);
> return ERR_PTR(ret);
> +
> +err_unlock:
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_unlock(attach->dmabuf->resv);
> +
> + dma_buf_detach(dmabuf, attach);
> + return ERR_PTR(ret);
> +}
> +EXPORT_SYMBOL_GPL(dma_buf_dynamic_attach);
> +
> +/**
> + * dma_buf_attach - Wrapper for dma_buf_dynamic_attach
> + * @dmabuf: [in] buffer to attach device to.
> + * @dev: [in] device to be attached.
> + *
> + * Wrapper to call dma_buf_dynamic_attach() for drivers which still use a static
> + * mapping.
> + */
> +struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf,
> + struct device *dev)
> +{
> + return dma_buf_dynamic_attach(dmabuf, dev, false);
> }
> EXPORT_SYMBOL_GPL(dma_buf_attach);
>
> @@ -711,11 +766,20 @@ void dma_buf_detach(struct dma_buf *dmabuf, struct dma_buf_attachment *attach)
> if (WARN_ON(!dmabuf || !attach))
> return;
>
> - if (attach->sgt)
> + if (attach->sgt) {
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_lock(attach->dmabuf->resv, NULL);
> +
> dmabuf->ops->unmap_dma_buf(attach, attach->sgt, attach->dir);
>
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_unlock(attach->dmabuf->resv);
> + }
> +
> mutex_lock(&dmabuf->lock);
> + dma_resv_lock(dmabuf->resv, NULL);
> list_del(&attach->node);
> + dma_resv_unlock(dmabuf->resv);
> if (dmabuf->ops->detach)
> dmabuf->ops->detach(dmabuf, attach);
>
> @@ -749,6 +813,9 @@ struct sg_table *dma_buf_map_attachment(struct dma_buf_attachment *attach,
> if (WARN_ON(!attach || !attach->dmabuf))
> return ERR_PTR(-EINVAL);
>
> + if (dma_buf_attachment_is_dynamic(attach))
> + dma_resv_assert_held(attach->dmabuf->resv);
> +
> if (attach->sgt) {
> /*
> * Two mappings with different directions for the same
> @@ -761,6 +828,9 @@ struct sg_table *dma_buf_map_attachment(struct dma_buf_attachment *attach,
> return attach->sgt;
> }
>
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_assert_held(attach->dmabuf->resv);
Almost tripped me up until I noticed your check for the buf/exporter here.
Nice check!
Maybe we could do an
else
lockdep_assert_not_held()
here? Open-coded ofc (or put it into drm_utils.h, I don't want to hold up
this patch any longer, then move it to lockdep.h later on), since
currently doesn't exist. Just to _really_ enforce the documented contract
here.
> +
> sg_table = attach->dmabuf->ops->map_dma_buf(attach, direction);
> if (!sg_table)
> sg_table = ERR_PTR(-ENOMEM);
> @@ -793,9 +863,15 @@ void dma_buf_unmap_attachment(struct dma_buf_attachment *attach,
> if (WARN_ON(!attach || !attach->dmabuf || !sg_table))
> return;
>
> + if (dma_buf_attachment_is_dynamic(attach))
> + dma_resv_assert_held(attach->dmabuf->resv);
> +
> if (attach->sgt == sg_table)
> return;
>
> + if (dma_buf_is_dynamic(attach->dmabuf))
> + dma_resv_assert_held(attach->dmabuf->resv);
> +
> attach->dmabuf->ops->unmap_dma_buf(attach, sg_table, direction);
> }
> EXPORT_SYMBOL_GPL(dma_buf_unmap_attachment);
> @@ -1219,10 +1295,12 @@ static int dma_buf_debug_show(struct seq_file *s, void *unused)
> seq_puts(s, "\tAttached Devices:\n");
> attach_count = 0;
>
> + dma_resv_lock(buf_obj->resv, NULL);
You've switched dmabuf->name to be protected by dma_resv too, so this
needs to be moved up above the seq_printf which prints ->name.
> list_for_each_entry(attach_obj, &buf_obj->attachments, node) {
> seq_printf(s, "\t%s\n", dev_name(attach_obj->dev));
> attach_count++;
> }
> + dma_resv_unlock(buf_obj->resv);
>
> seq_printf(s, "Total %d devices attached\n\n",
> attach_count);
> diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h
> index ec212cb27fdc..bcc0f4d0b678 100644
> --- a/include/linux/dma-buf.h
> +++ b/include/linux/dma-buf.h
> @@ -42,6 +42,18 @@ struct dma_buf_ops {
> */
> bool cache_sgt_mapping;
>
> + /**
> + * @dynamic_mapping:
> + *
> + * If true the framework makes sure that the map/unmap_dma_buf
> + * callbacks are always called with the dma_resv object locked.
> + *
> + * If false the framework makes ure that the map/unmap_dma_buf
s/ure/sure/
> + * callbacks are always called without the dma_resv object locked.
> + * Mutual exclusive with @cache_sgt_mapping.
> + */
> + bool dynamic_mapping;
> +
> /**
> * @attach:
> *
> @@ -109,6 +121,9 @@ struct dma_buf_ops {
> * any other kind of sharing that the exporter might wish to make
> * available to buffer-users.
> *
> + * This is always called with the dmabuf->resv object locked when
> + * the dynamic_mapping flag is true.
> + *
> * Returns:
> *
> * A &sg_table scatter list of or the backing storage of the DMA buffer,
> @@ -267,7 +282,8 @@ struct dma_buf_ops {
> * struct dma_buf - shared buffer object
> * @size: size of the buffer
> * @file: file pointer used for sharing buffers across, and for refcounting.
> - * @attachments: list of dma_buf_attachment that denotes all devices attached.
> + * @attachments: list of dma_buf_attachment that denotes all devices attached,
> + * protected by dma_resv lock.
> * @ops: dma_buf_ops associated with this buffer object.
> * @lock: used internally to serialize list manipulation, attach/detach and
> * vmap/unmap, and accesses to name
@name in dma_buf needs to be updated to mention that it's protected by
dma_resv lock.
> @@ -323,10 +339,12 @@ struct dma_buf {
> * struct dma_buf_attachment - holds device-buffer attachment data
> * @dmabuf: buffer for this attachment.
> * @dev: device attached to the buffer.
> - * @node: list of dma_buf_attachment.
> + * @node: list of dma_buf_attachment, protected by dma_resv lock of the dmabuf.
> * @sgt: cached mapping.
> * @dir: direction of cached mapping.
> * @priv: exporter specific attachment data.
> + * @dynamic_mapping: true if dma_buf_map/unmap_attachment() is called with the
> + * dma_resv lock held.
> *
> * This structure holds the attachment information between the dma_buf buffer
> * and its user device(s). The list contains one attachment struct per device
> @@ -343,6 +361,7 @@ struct dma_buf_attachment {
> struct list_head node;
> struct sg_table *sgt;
> enum dma_data_direction dir;
> + bool dynamic_mapping;
> void *priv;
> };
>
> @@ -394,10 +413,39 @@ static inline void get_dma_buf(struct dma_buf *dmabuf)
> get_file(dmabuf->file);
> }
>
> +/**
> + * dma_buf_is_dynamic - check if a DMA-buf uses dynamic mappings.
> + * @dmabuf: the DMA-buf to check
> + *
> + * Returns true if a DMA-buf exporter wants to be called with the dma_resv
> + * locked, false if it doesn't wants to be called with the lock held.
Maybe added "... for the map/unmap callbacks, ..." since we don't
guarantee to hold the lock for all callbacks. You clarify this already
below.
> + */
> +static inline bool dma_buf_is_dynamic(struct dma_buf *dmabuf)
> +{
> + return dmabuf->ops->dynamic_mapping;
> +}
> +
> +/**
> + * dma_buf_attachment_is_dynamic - check if a DMA-buf attachment uses dynamic
> + * mappinsg
> + * @attach: the DMA-buf attachment to check
> + *
> + * Returns true if a DMA-buf importer wants to call the map/unmap functions with
> + * the dma_resv lock held.
> + */
> +static inline bool
> +dma_buf_attachment_is_dynamic(struct dma_buf_attachment *attach)
> +{
> + return attach->dynamic_mapping;
> +}
> +
> struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf,
> - struct device *dev);
> + struct device *dev);
> +struct dma_buf_attachment *
> +dma_buf_dynamic_attach(struct dma_buf *dmabuf, struct device *dev,
> + bool dynamic_mapping);
> void dma_buf_detach(struct dma_buf *dmabuf,
> - struct dma_buf_attachment *dmabuf_attach);
> + struct dma_buf_attachment *attach);
>
> struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
>
> @@ -409,6 +457,7 @@ struct sg_table *dma_buf_map_attachment(struct dma_buf_attachment *,
> enum dma_data_direction);
> void dma_buf_unmap_attachment(struct dma_buf_attachment *, struct sg_table *,
> enum dma_data_direction);
> +void dma_buf_move_notify(struct dma_buf *dma_buf);
> int dma_buf_begin_cpu_access(struct dma_buf *dma_buf,
> enum dma_data_direction dir);
> int dma_buf_end_cpu_access(struct dma_buf *dma_buf,
With the nits all addressed:
Reviewed-by: Daniel Vetter <daniel.vetter@xxxxxxxx>
> --
> 2.17.1
>
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
