Re: [PATCH 3/8] xen/grant-table: Allow allocating buffers suitable for DMA

From: Boris Ostrovsky
Date: Tue May 29 2018 - 15:07:25 EST


On 05/25/2018 11:33 AM, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@xxxxxxxx>
>
> Extend grant table module API to allow allocating buffers that can
> be used for DMA operations and mapping foreign grant references
> on top of those.
> The resulting buffer is similar to the one allocated by the balloon
> driver in terms that proper memory reservation is made
> ({increase|decrease}_reservation and VA mappings updated if needed).
> This is useful for sharing foreign buffers with HW drivers which
> cannot work with scattered buffers provided by the balloon driver,
> but require DMAable memory instead.
>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@xxxxxxxx>
> ---
> drivers/xen/Kconfig | 13 ++++
> drivers/xen/grant-table.c | 124 ++++++++++++++++++++++++++++++++++++++
> include/xen/grant_table.h | 25 ++++++++
> 3 files changed, 162 insertions(+)
>
> diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig
> index e5d0c28372ea..3431fe210624 100644
> --- a/drivers/xen/Kconfig
> +++ b/drivers/xen/Kconfig
> @@ -161,6 +161,19 @@ config XEN_GRANT_DEV_ALLOC
> to other domains. This can be used to implement frontend drivers
> or as part of an inter-domain shared memory channel.
>
> +config XEN_GRANT_DMA_ALLOC
> + bool "Allow allocating DMA capable buffers with grant reference module"
> + depends on XEN


Should it depend on anything from DMA? CONFIG_HAS_DMA for example?

> + help
> + Extends grant table module API to allow allocating DMA capable
> + buffers and mapping foreign grant references on top of it.
> + The resulting buffer is similar to one allocated by the balloon
> + driver in terms that proper memory reservation is made
> + ({increase|decrease}_reservation and VA mappings updated if needed).
> + This is useful for sharing foreign buffers with HW drivers which
> + cannot work with scattered buffers provided by the balloon driver,
> + but require DMAable memory instead.
> +
> config SWIOTLB_XEN
> def_bool y
> select SWIOTLB
> diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c
> index d7488226e1f2..06fe6e7f639c 100644
> --- a/drivers/xen/grant-table.c
> +++ b/drivers/xen/grant-table.c
> @@ -45,6 +45,9 @@
> #include <linux/workqueue.h>
> #include <linux/ratelimit.h>
> #include <linux/moduleparam.h>
> +#ifdef CONFIG_XEN_GRANT_DMA_ALLOC
> +#include <linux/dma-mapping.h>
> +#endif
>
> #include <xen/xen.h>
> #include <xen/interface/xen.h>
> @@ -57,6 +60,7 @@
> #ifdef CONFIG_X86
> #include <asm/xen/cpuid.h>
> #endif
> +#include <xen/mem_reservation.h>
> #include <asm/xen/hypercall.h>
> #include <asm/xen/interface.h>
>
> @@ -811,6 +815,82 @@ int gnttab_alloc_pages(int nr_pages, struct page **pages)
> }
> EXPORT_SYMBOL(gnttab_alloc_pages);
>
> +#ifdef CONFIG_XEN_GRANT_DMA_ALLOC
> +/**
> + * gnttab_dma_alloc_pages - alloc DMAable pages suitable for grant mapping into
> + * @args: arguments to the function
> + */
> +int gnttab_dma_alloc_pages(struct gnttab_dma_alloc_args *args)
> +{
> + unsigned long pfn, start_pfn;
> + xen_pfn_t *frames;
> + size_t size;
> + int i, ret;
> +
> + frames = kcalloc(args->nr_pages, sizeof(*frames), GFP_KERNEL);
> + if (!frames)
> + return -ENOMEM;
> +
> + size = args->nr_pages << PAGE_SHIFT;
> + if (args->coherent)
> + args->vaddr = dma_alloc_coherent(args->dev, size,
> + &args->dev_bus_addr,
> + GFP_KERNEL | __GFP_NOWARN);
> + else
> + args->vaddr = dma_alloc_wc(args->dev, size,
> + &args->dev_bus_addr,
> + GFP_KERNEL | __GFP_NOWARN);
> + if (!args->vaddr) {
> + pr_err("Failed to allocate DMA buffer of size %zu\n", size);
> + ret = -ENOMEM;
> + goto fail_free_frames;
> + }
> +
> + start_pfn = __phys_to_pfn(args->dev_bus_addr);
> + for (pfn = start_pfn, i = 0; pfn < start_pfn + args->nr_pages;
> + pfn++, i++) {
> + struct page *page = pfn_to_page(pfn);
> +
> + args->pages[i] = page;
> + frames[i] = xen_page_to_gfn(page);
> + xenmem_reservation_scrub_page(page);
> + }
> +
> + xenmem_reservation_va_mapping_reset(args->nr_pages, args->pages);
> +
> + ret = xenmem_reservation_decrease(args->nr_pages, frames);
> + if (ret != args->nr_pages) {
> + pr_err("Failed to decrease reservation for DMA buffer\n");
> + xenmem_reservation_increase(ret, frames);
> + ret = -EFAULT;
> + goto fail_free_dma;
> + }
> +
> + ret = gnttab_pages_set_private(args->nr_pages, args->pages);
> + if (ret < 0)
> + goto fail_clear_private;
> +
> + kfree(frames);
> + return 0;
> +
> +fail_clear_private:
> + gnttab_pages_clear_private(args->nr_pages, args->pages);
> +fail_free_dma:


Do you need to xenmem_reservation_increase()?

> + xenmem_reservation_va_mapping_update(args->nr_pages, args->pages,
> + frames);
> + if (args->coherent)
> + dma_free_coherent(args->dev, size,
> + args->vaddr, args->dev_bus_addr);
> + else
> + dma_free_wc(args->dev, size,
> + args->vaddr, args->dev_bus_addr);
> +fail_free_frames:
> + kfree(frames);
> + return ret;
> +}
> +EXPORT_SYMBOL(gnttab_dma_alloc_pages);
> +#endif
> +
> void gnttab_pages_clear_private(int nr_pages, struct page **pages)
> {
> int i;
> @@ -838,6 +918,50 @@ void gnttab_free_pages(int nr_pages, struct page **pages)
> }
> EXPORT_SYMBOL(gnttab_free_pages);
>
> +#ifdef CONFIG_XEN_GRANT_DMA_ALLOC
> +/**
> + * gnttab_dma_free_pages - free DMAable pages
> + * @args: arguments to the function
> + */
> +int gnttab_dma_free_pages(struct gnttab_dma_alloc_args *args)
> +{
> + xen_pfn_t *frames;
> + size_t size;
> + int i, ret;
> +
> + gnttab_pages_clear_private(args->nr_pages, args->pages);
> +
> + frames = kcalloc(args->nr_pages, sizeof(*frames), GFP_KERNEL);


Any way you can do it without allocating memory? One possibility is to
keep allocated frames from gnttab_dma_alloc_pages(). (Not sure I like
that either but it's the only thing I can think of).


> + if (!frames)
> + return -ENOMEM;
> +
> + for (i = 0; i < args->nr_pages; i++)
> + frames[i] = page_to_xen_pfn(args->pages[i]);


Not xen_page_to_gfn()?


> +
> + ret = xenmem_reservation_increase(args->nr_pages, frames);
> + if (ret != args->nr_pages) {
> + pr_err("Failed to decrease reservation for DMA buffer\n");
> + ret = -EFAULT;
> + } else {
> + ret = 0;
> + }
> +
> + xenmem_reservation_va_mapping_update(args->nr_pages, args->pages,
> + frames);
> +
> + size = args->nr_pages << PAGE_SHIFT;
> + if (args->coherent)
> + dma_free_coherent(args->dev, size,
> + args->vaddr, args->dev_bus_addr);
> + else
> + dma_free_wc(args->dev, size,
> + args->vaddr, args->dev_bus_addr);
> + kfree(frames);
> + return ret;
> +}
> +EXPORT_SYMBOL(gnttab_dma_free_pages);
> +#endif
> +
> /* Handling of paged out grant targets (GNTST_eagain) */
> #define MAX_DELAY 256
> static inline void
> diff --git a/include/xen/grant_table.h b/include/xen/grant_table.h
> index de03f2542bb7..982e34242b9c 100644
> --- a/include/xen/grant_table.h
> +++ b/include/xen/grant_table.h
> @@ -198,6 +198,31 @@ void gnttab_free_auto_xlat_frames(void);
> int gnttab_alloc_pages(int nr_pages, struct page **pages);
> void gnttab_free_pages(int nr_pages, struct page **pages);
>
> +#ifdef CONFIG_XEN_GRANT_DMA_ALLOC
> +struct gnttab_dma_alloc_args {
> + /* Device for which DMA memory will be/was allocated. */
> + struct device *dev;
> + /*
> + * If set then DMA buffer is coherent and write-combine otherwise.
> + */

Single-line comment

> + bool coherent;
> + /*
> + * Number of entries in the @pages array, defines the size
> + * of the DMA buffer.
> + */

This can be made into single line comment as well. However, I am not
sure this comment as well as those below are necessary. Fields names are
self-describing IMO.


-boris

> + int nr_pages;
> + /* Array of pages @pages filled with pages of the DMA buffer. */
> + struct page **pages;
> + /* Virtual/CPU address of the DMA buffer. */
> + void *vaddr;
> + /* Bus address of the DMA buffer. */
> + dma_addr_t dev_bus_addr;
> +};
> +
> +int gnttab_dma_alloc_pages(struct gnttab_dma_alloc_args *args);
> +int gnttab_dma_free_pages(struct gnttab_dma_alloc_args *args);
> +#endif
> +
> int gnttab_pages_set_private(int nr_pages, struct page **pages);
> void gnttab_pages_clear_private(int nr_pages, struct page **pages);
>