Message ID | 20181115154912.GA27969@jordon-HP-15-Notebook-PC (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | [1/9] mm: Introduce new vm_insert_range API | expand |
On 11/15/18 5:49 PM, Souptick Joarder wrote: > Convert to use vm_insert_range() to map range of kernel > memory to user vma. > > Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com> > Reviewed-by: Matthew Wilcox <willy@infradead.org> > --- > drivers/gpu/drm/xen/xen_drm_front_gem.c | 20 ++++++-------------- > 1 file changed, 6 insertions(+), 14 deletions(-) > > diff --git a/drivers/gpu/drm/xen/xen_drm_front_gem.c b/drivers/gpu/drm/xen/xen_drm_front_gem.c > index 47ff019..a3eade6 100644 > --- a/drivers/gpu/drm/xen/xen_drm_front_gem.c > +++ b/drivers/gpu/drm/xen/xen_drm_front_gem.c > @@ -225,8 +225,7 @@ struct drm_gem_object * > static int gem_mmap_obj(struct xen_gem_object *xen_obj, > struct vm_area_struct *vma) > { > - unsigned long addr = vma->vm_start; > - int i; > + int err; I would love to keep ret, not err > > /* > * clear the VM_PFNMAP flag that was set by drm_gem_mmap(), and set the > @@ -247,18 +246,11 @@ static int gem_mmap_obj(struct xen_gem_object *xen_obj, > * FIXME: as we insert all the pages now then no .fault handler must > * be called, so don't provide one > */ > - for (i = 0; i < xen_obj->num_pages; i++) { > - int ret; > - > - ret = vm_insert_page(vma, addr, xen_obj->pages[i]); > - if (ret < 0) { > - DRM_ERROR("Failed to insert pages into vma: %d\n", ret); > - return ret; > - } > - > - addr += PAGE_SIZE; > - } > - return 0; > + err = vm_insert_range(vma, vma->vm_start, xen_obj->pages, > + xen_obj->num_pages); > + if (err < 0) > + DRM_ERROR("Failed to insert pages into vma: %d\n", err); > + return err; > } > > int xen_drm_front_gem_mmap(struct file *filp, struct vm_area_struct *vma) With the above fixed, Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
On Mon, Nov 19, 2018 at 3:22 PM Oleksandr Andrushchenko <andr2000@gmail.com> wrote: > > On 11/15/18 5:49 PM, Souptick Joarder wrote: > > Convert to use vm_insert_range() to map range of kernel > > memory to user vma. > > > > Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com> > > Reviewed-by: Matthew Wilcox <willy@infradead.org> > > --- > > drivers/gpu/drm/xen/xen_drm_front_gem.c | 20 ++++++-------------- > > 1 file changed, 6 insertions(+), 14 deletions(-) > > > > diff --git a/drivers/gpu/drm/xen/xen_drm_front_gem.c b/drivers/gpu/drm/xen/xen_drm_front_gem.c > > index 47ff019..a3eade6 100644 > > --- a/drivers/gpu/drm/xen/xen_drm_front_gem.c > > +++ b/drivers/gpu/drm/xen/xen_drm_front_gem.c > > @@ -225,8 +225,7 @@ struct drm_gem_object * > > static int gem_mmap_obj(struct xen_gem_object *xen_obj, > > struct vm_area_struct *vma) > > { > > - unsigned long addr = vma->vm_start; > > - int i; > > + int err; > I would love to keep ret, not err Sure, will add it in v2. But I think, err is more appropriate here. > > > > /* > > * clear the VM_PFNMAP flag that was set by drm_gem_mmap(), and set the > > @@ -247,18 +246,11 @@ static int gem_mmap_obj(struct xen_gem_object *xen_obj, > > * FIXME: as we insert all the pages now then no .fault handler must > > * be called, so don't provide one > > */ > > - for (i = 0; i < xen_obj->num_pages; i++) { > > - int ret; > > - > > - ret = vm_insert_page(vma, addr, xen_obj->pages[i]); > > - if (ret < 0) { > > - DRM_ERROR("Failed to insert pages into vma: %d\n", ret); > > - return ret; > > - } > > - > > - addr += PAGE_SIZE; > > - } > > - return 0; > > + err = vm_insert_range(vma, vma->vm_start, xen_obj->pages, > > + xen_obj->num_pages); > > + if (err < 0) > > + DRM_ERROR("Failed to insert pages into vma: %d\n", err); > > + return err; > > } > > > > int xen_drm_front_gem_mmap(struct file *filp, struct vm_area_struct *vma) > > With the above fixed, > > Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com> >
On 11/19/18 12:42 PM, Souptick Joarder wrote: > On Mon, Nov 19, 2018 at 3:22 PM Oleksandr Andrushchenko > <andr2000@gmail.com> wrote: >> On 11/15/18 5:49 PM, Souptick Joarder wrote: >>> Convert to use vm_insert_range() to map range of kernel >>> memory to user vma. >>> >>> Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com> >>> Reviewed-by: Matthew Wilcox <willy@infradead.org> >>> --- >>> drivers/gpu/drm/xen/xen_drm_front_gem.c | 20 ++++++-------------- >>> 1 file changed, 6 insertions(+), 14 deletions(-) >>> >>> diff --git a/drivers/gpu/drm/xen/xen_drm_front_gem.c b/drivers/gpu/drm/xen/xen_drm_front_gem.c >>> index 47ff019..a3eade6 100644 >>> --- a/drivers/gpu/drm/xen/xen_drm_front_gem.c >>> +++ b/drivers/gpu/drm/xen/xen_drm_front_gem.c >>> @@ -225,8 +225,7 @@ struct drm_gem_object * >>> static int gem_mmap_obj(struct xen_gem_object *xen_obj, >>> struct vm_area_struct *vma) >>> { >>> - unsigned long addr = vma->vm_start; >>> - int i; >>> + int err; >> I would love to keep ret, not err > Sure, will add it in v2. > But I think, err is more appropriate here. I used "ret" throughout the driver, so this is just to remain consistent: grep -rnw err drivers/gpu/drm/xen/ | wc -l 0 grep -rnw ret drivers/gpu/drm/xen/ | wc -l 204 >>> /* >>> * clear the VM_PFNMAP flag that was set by drm_gem_mmap(), and set the >>> @@ -247,18 +246,11 @@ static int gem_mmap_obj(struct xen_gem_object *xen_obj, >>> * FIXME: as we insert all the pages now then no .fault handler must >>> * be called, so don't provide one >>> */ >>> - for (i = 0; i < xen_obj->num_pages; i++) { >>> - int ret; >>> - >>> - ret = vm_insert_page(vma, addr, xen_obj->pages[i]); >>> - if (ret < 0) { >>> - DRM_ERROR("Failed to insert pages into vma: %d\n", ret); >>> - return ret; >>> - } >>> - >>> - addr += PAGE_SIZE; >>> - } >>> - return 0; >>> + err = vm_insert_range(vma, vma->vm_start, xen_obj->pages, >>> + xen_obj->num_pages); >>> + if (err < 0) >>> + DRM_ERROR("Failed to insert pages into vma: %d\n", err); >>> + return err; >>> } >>> >>> int xen_drm_front_gem_mmap(struct file *filp, struct vm_area_struct *vma) >> With the above fixed, >> >> Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com> >>
On Mon, Nov 19, 2018 at 01:02:46PM +0200, Oleksandr Andrushchenko wrote: > On 11/19/18 12:42 PM, Souptick Joarder wrote: > > On Mon, Nov 19, 2018 at 3:22 PM Oleksandr Andrushchenko > > <andr2000@gmail.com> wrote: > > > > - unsigned long addr = vma->vm_start; > > > > - int i; > > > > + int err; > > > I would love to keep ret, not err > > Sure, will add it in v2. > > But I think, err is more appropriate here. > > I used "ret" throughout the driver, so this is just to remain consistent: > > grep -rnw err drivers/gpu/drm/xen/ | wc -l > 0 > grep -rnw ret drivers/gpu/drm/xen/ | wc -l > 204 It's your driver, so that's fine. The reason we chose 'err' over 'ret' is that there's a history of errno vs VM_FAULT_xxx code confusion in this area. Naming a variable 'err' makes it clear this is an errno and not a vm_fault_t. > > > With the above fixed, > > > > > > Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com> Thanks.
diff --git a/drivers/gpu/drm/xen/xen_drm_front_gem.c b/drivers/gpu/drm/xen/xen_drm_front_gem.c index 47ff019..a3eade6 100644 --- a/drivers/gpu/drm/xen/xen_drm_front_gem.c +++ b/drivers/gpu/drm/xen/xen_drm_front_gem.c @@ -225,8 +225,7 @@ struct drm_gem_object * static int gem_mmap_obj(struct xen_gem_object *xen_obj, struct vm_area_struct *vma) { - unsigned long addr = vma->vm_start; - int i; + int err; /* * clear the VM_PFNMAP flag that was set by drm_gem_mmap(), and set the @@ -247,18 +246,11 @@ static int gem_mmap_obj(struct xen_gem_object *xen_obj, * FIXME: as we insert all the pages now then no .fault handler must * be called, so don't provide one */ - for (i = 0; i < xen_obj->num_pages; i++) { - int ret; - - ret = vm_insert_page(vma, addr, xen_obj->pages[i]); - if (ret < 0) { - DRM_ERROR("Failed to insert pages into vma: %d\n", ret); - return ret; - } - - addr += PAGE_SIZE; - } - return 0; + err = vm_insert_range(vma, vma->vm_start, xen_obj->pages, + xen_obj->num_pages); + if (err < 0) + DRM_ERROR("Failed to insert pages into vma: %d\n", err); + return err; } int xen_drm_front_gem_mmap(struct file *filp, struct vm_area_struct *vma)