From: Chris Wilson <chris@chris-wilson.co.uk>
To: dri-devel@lists.freedesktop.org
Cc: intel-gfx@lists.freedesktop.org
Subject: [PATCH v3 3/3] drm/vgem: Keep a reference to the dmabuf when mmaped
Date: Fri, 30 Sep 2016 14:59:59 +0100 [thread overview]
Message-ID: <20160930135959.9159-3-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20160930135959.9159-1-chris@chris-wilson.co.uk>
In order to keep the dmabuf alive whilst the mmap is, we need to hold a
reference to the dmabuf and not the backing object. This is important as
the dmabuf not only keeps the object alive, but also the device so that
dmabuf = vgem_create_dmabuf();
ptr = mmap(... dmabuf ...);
close(dmabuf);
persists across module-unload as well as device closure.
Testcase: igt/vgem_basic/unload
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Petri Latvala <petri.latvala@intel.com>
---
drivers/gpu/drm/vgem/vgem_drv.c | 17 +++--------------
1 file changed, 3 insertions(+), 14 deletions(-)
diff --git a/drivers/gpu/drm/vgem/vgem_drv.c b/drivers/gpu/drm/vgem/vgem_drv.c
index f36c14729b55..74a83e41efa9 100644
--- a/drivers/gpu/drm/vgem/vgem_drv.c
+++ b/drivers/gpu/drm/vgem/vgem_drv.c
@@ -198,7 +198,6 @@ static struct drm_ioctl_desc vgem_ioctls[] = {
static int vgem_mmap(struct file *filp, struct vm_area_struct *vma)
{
- unsigned long flags = vma->vm_flags;
int ret;
ret = drm_gem_mmap(filp, vma);
@@ -208,7 +207,7 @@ static int vgem_mmap(struct file *filp, struct vm_area_struct *vma)
/* Keep the WC mmaping set by drm_gem_mmap() but our pages
* are ordinary and not special.
*/
- vma->vm_flags = flags | VM_DONTEXPAND | VM_DONTDUMP;
+ vma->vm_flags &= ~(VM_IO | VM_PFNMAP);
return 0;
}
@@ -281,21 +280,11 @@ static int vgem_prime_mmap(struct drm_gem_object *obj,
{
int ret;
- if (obj->size < vma->vm_end - vma->vm_start)
- return -EINVAL;
-
- if (!obj->filp)
- return -ENODEV;
-
- ret = obj->filp->f_op->mmap(obj->filp, vma);
+ ret = drm_gem_mmap_obj(obj, obj->size, vma);
if (ret)
return ret;
- fput(vma->vm_file);
- vma->vm_file = get_file(obj->filp);
- vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP;
- vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
-
+ vma->vm_flags &= ~(VM_IO | VM_PFNMAP);
return 0;
}
--
2.9.3
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2016-09-30 13:59 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-30 13:59 [PATCH v3 1/3] drm/prime: Pass the right module owner through to dma_buf_export() Chris Wilson
2016-09-30 13:59 ` [PATCH v3 2/3] drm/prime: Take a ref on the drm_dev when exporting a dma_buf Chris Wilson
2016-09-30 13:59 ` Chris Wilson [this message]
2016-10-05 13:11 ` [PATCH v3 3/3] drm/vgem: Keep a reference to the dmabuf when mmaped Daniel Vetter
2016-10-05 13:40 ` Chris Wilson
2016-10-05 13:42 ` Daniel Vetter
2016-09-30 14:22 ` [PATCH v3 1/3] drm/prime: Pass the right module owner through to dma_buf_export() Christian König
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160930135959.9159-3-chris@chris-wilson.co.uk \
--to=chris@chris-wilson.co.uk \
--cc=dri-devel@lists.freedesktop.org \
--cc=intel-gfx@lists.freedesktop.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).