* [PATCH v2] gpu: drm: gma500: Change return type to vm_fault_t
@ 2018-05-23 14:46 Souptick Joarder
2018-05-26 2:41 ` kbuild test robot
0 siblings, 1 reply; 3+ messages in thread
From: Souptick Joarder @ 2018-05-23 14:46 UTC (permalink / raw)
To: patrik.r.jakobsson, airlied; +Cc: dri-devel, linux-kernel, willy
Use new return type vm_fault_t for fault handler. For
now, this is just documenting that the function returns
a VM_FAULT value rather than an errno. Once all instances
are converted, vm_fault_t will become a distinct type.
Ref-> commit 1c8f422059ae ("mm: change return type to vm_fault_t")
Previously vm_insert_{pfn,mixed} returns err which driver
mapped into VM_FAULT_* type. The new function
vmf_insert_{pfn,mixed} will replace this inefficiency by
returning VM_FAULT_* type.
vmf_error() is the newly introduce inline function
in 4.17-rc6.
Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com>
Reviewed-by: Matthew Wilcox <mawilcox@microsoft.com>
---
v2: updated the change log
drivers/gpu/drm/gma500/framebuffer.c | 14 +++++---------
drivers/gpu/drm/gma500/gem.c | 27 ++++++++++-----------------
drivers/gpu/drm/gma500/psb_drv.h | 3 ++-
3 files changed, 17 insertions(+), 27 deletions(-)
diff --git a/drivers/gpu/drm/gma500/framebuffer.c b/drivers/gpu/drm/gma500/framebuffer.c
index cb0a2ae..7635ce8 100644
--- a/drivers/gpu/drm/gma500/framebuffer.c
+++ b/drivers/gpu/drm/gma500/framebuffer.c
@@ -111,7 +111,7 @@ static int psbfb_pan(struct fb_var_screeninfo *var, struct fb_info *info)
return 0;
}
-static int psbfb_vm_fault(struct vm_fault *vmf)
+static vm_fault_t psbfb_vm_fault(struct vm_fault *vmf)
{
struct vm_area_struct *vma = vmf->vma;
struct psb_framebuffer *psbfb = vma->vm_private_data;
@@ -120,7 +120,7 @@ static int psbfb_vm_fault(struct vm_fault *vmf)
int page_num;
int i;
unsigned long address;
- int ret;
+ vm_fault_t ret;
unsigned long pfn;
unsigned long phys_addr = (unsigned long)dev_priv->stolen_base +
psbfb->gtt->offset;
@@ -133,18 +133,14 @@ static int psbfb_vm_fault(struct vm_fault *vmf)
for (i = 0; i < page_num; i++) {
pfn = (phys_addr >> PAGE_SHIFT);
- ret = vm_insert_mixed(vma, address,
+ ret = vmf_insert_mixed(vma, address,
__pfn_to_pfn_t(pfn, PFN_DEV));
- if (unlikely((ret == -EBUSY) || (ret != 0 && i > 0)))
+ if (unlikely(ret & VM_FAULT_ERROR))
break;
- else if (unlikely(ret != 0)) {
- ret = (ret == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS;
- return ret;
- }
address += PAGE_SIZE;
phys_addr += PAGE_SIZE;
}
- return VM_FAULT_NOPAGE;
+ return ret;
}
static void psbfb_vm_open(struct vm_area_struct *vma)
diff --git a/drivers/gpu/drm/gma500/gem.c b/drivers/gpu/drm/gma500/gem.c
index 1312397..e7be5c9 100644
--- a/drivers/gpu/drm/gma500/gem.c
+++ b/drivers/gpu/drm/gma500/gem.c
@@ -134,12 +134,13 @@ int psb_gem_dumb_create(struct drm_file *file, struct drm_device *dev,
* vma->vm_private_data points to the GEM object that is backing this
* mapping.
*/
-int psb_gem_fault(struct vm_fault *vmf)
+vm_fault_t psb_gem_fault(struct vm_fault *vmf)
{
struct vm_area_struct *vma = vmf->vma;
struct drm_gem_object *obj;
struct gtt_range *r;
- int ret;
+ int err;
+ vm_fault_t ret;
unsigned long pfn;
pgoff_t page_offset;
struct drm_device *dev;
@@ -158,9 +159,10 @@ int psb_gem_fault(struct vm_fault *vmf)
/* For now the mmap pins the object and it stays pinned. As things
stand that will do us no harm */
if (r->mmapping == 0) {
- ret = psb_gtt_pin(r);
- if (ret < 0) {
- dev_err(dev->dev, "gma500: pin failed: %d\n", ret);
+ err = psb_gtt_pin(r);
+ if (err < 0) {
+ dev_err(dev->dev, "gma500: pin failed: %d\n", err);
+ ret = vmf_error(err);
goto fail;
}
r->mmapping = 1;
@@ -175,18 +177,9 @@ int psb_gem_fault(struct vm_fault *vmf)
pfn = (dev_priv->stolen_base + r->offset) >> PAGE_SHIFT;
else
pfn = page_to_pfn(r->pages[page_offset]);
- ret = vm_insert_pfn(vma, vmf->address, pfn);
-
+ ret = vmf_insert_pfn(vma, vmf->address, pfn);
fail:
mutex_unlock(&dev_priv->mmap_mutex);
- switch (ret) {
- case 0:
- case -ERESTARTSYS:
- case -EINTR:
- return VM_FAULT_NOPAGE;
- case -ENOMEM:
- return VM_FAULT_OOM;
- default:
- return VM_FAULT_SIGBUS;
- }
+
+ return ret;
}
diff --git a/drivers/gpu/drm/gma500/psb_drv.h b/drivers/gpu/drm/gma500/psb_drv.h
index e8300f5..93d2f40 100644
--- a/drivers/gpu/drm/gma500/psb_drv.h
+++ b/drivers/gpu/drm/gma500/psb_drv.h
@@ -21,6 +21,7 @@
#define _PSB_DRV_H_
#include <linux/kref.h>
+#include <linux/mm_types.h>
#include <drm/drmP.h>
#include <drm/drm_global.h>
@@ -749,7 +750,7 @@ extern int psb_gem_get_aperture(struct drm_device *dev, void *data,
struct drm_file *file);
extern int psb_gem_dumb_create(struct drm_file *file, struct drm_device *dev,
struct drm_mode_create_dumb *args);
-extern int psb_gem_fault(struct vm_fault *vmf);
+extern vm_fault_t psb_gem_fault(struct vm_fault *vmf);
/* psb_device.c */
extern const struct psb_ops psb_chip_ops;
--
1.9.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH v2] gpu: drm: gma500: Change return type to vm_fault_t
2018-05-23 14:46 [PATCH v2] gpu: drm: gma500: Change return type to vm_fault_t Souptick Joarder
@ 2018-05-26 2:41 ` kbuild test robot
2018-05-26 13:32 ` Souptick Joarder
0 siblings, 1 reply; 3+ messages in thread
From: kbuild test robot @ 2018-05-26 2:41 UTC (permalink / raw)
To: Souptick Joarder
Cc: kbuild-all, patrik.r.jakobsson, airlied, dri-devel, linux-kernel, willy
[-- Attachment #1: Type: text/plain, Size: 2325 bytes --]
Hi Souptick,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on drm/drm-next]
[also build test WARNING on v4.17-rc6]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
url: https://github.com/0day-ci/linux/commits/Souptick-Joarder/gpu-drm-gma500-Change-return-type-to-vm_fault_t/20180526-084629
base: git://people.freedesktop.org/~airlied/linux.git drm-next
config: x86_64-randconfig-x013-201820 (attached as .config)
compiler: gcc-7 (Debian 7.3.0-16) 7.3.0
reproduce:
# save the attached .config to linux build tree
make ARCH=x86_64
Note: it may well be a FALSE warning. FWIW you are at least aware of it now.
http://gcc.gnu.org/wiki/Better_Uninitialized_Warnings
All warnings (new ones prefixed by >>):
drivers/gpu//drm/gma500/framebuffer.c: In function 'psbfb_vm_fault':
>> drivers/gpu//drm/gma500/framebuffer.c:143:9: warning: 'ret' may be used uninitialized in this function [-Wmaybe-uninitialized]
return ret;
^~~
vim +/ret +143 drivers/gpu//drm/gma500/framebuffer.c
113
114 static vm_fault_t psbfb_vm_fault(struct vm_fault *vmf)
115 {
116 struct vm_area_struct *vma = vmf->vma;
117 struct psb_framebuffer *psbfb = vma->vm_private_data;
118 struct drm_device *dev = psbfb->base.dev;
119 struct drm_psb_private *dev_priv = dev->dev_private;
120 int page_num;
121 int i;
122 unsigned long address;
123 vm_fault_t ret;
124 unsigned long pfn;
125 unsigned long phys_addr = (unsigned long)dev_priv->stolen_base +
126 psbfb->gtt->offset;
127
128 page_num = vma_pages(vma);
129 address = vmf->address - (vmf->pgoff << PAGE_SHIFT);
130
131 vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
132
133 for (i = 0; i < page_num; i++) {
134 pfn = (phys_addr >> PAGE_SHIFT);
135
136 ret = vmf_insert_mixed(vma, address,
137 __pfn_to_pfn_t(pfn, PFN_DEV));
138 if (unlikely(ret & VM_FAULT_ERROR))
139 break;
140 address += PAGE_SIZE;
141 phys_addr += PAGE_SIZE;
142 }
> 143 return ret;
144 }
145
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 33960 bytes --]
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH v2] gpu: drm: gma500: Change return type to vm_fault_t
2018-05-26 2:41 ` kbuild test robot
@ 2018-05-26 13:32 ` Souptick Joarder
0 siblings, 0 replies; 3+ messages in thread
From: Souptick Joarder @ 2018-05-26 13:32 UTC (permalink / raw)
To: kbuild test robot
Cc: kbuild-all, Patrik Jakobsson, airlied, dri-devel, linux-kernel,
Matthew Wilcox
On Sat, May 26, 2018 at 8:11 AM, kbuild test robot <lkp@intel.com> wrote:
> Hi Souptick,
>
> Thank you for the patch! Perhaps something to improve:
>
> [auto build test WARNING on drm/drm-next]
> [also build test WARNING on v4.17-rc6]
> [if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
>
> url: https://github.com/0day-ci/linux/commits/Souptick-Joarder/gpu-drm-gma500-Change-return-type-to-vm_fault_t/20180526-084629
> base: git://people.freedesktop.org/~airlied/linux.git drm-next
> config: x86_64-randconfig-x013-201820 (attached as .config)
> compiler: gcc-7 (Debian 7.3.0-16) 7.3.0
> reproduce:
> # save the attached .config to linux build tree
> make ARCH=x86_64
>
> Note: it may well be a FALSE warning. FWIW you are at least aware of it now.
> http://gcc.gnu.org/wiki/Better_Uninitialized_Warnings
>
> All warnings (new ones prefixed by >>):
>
> drivers/gpu//drm/gma500/framebuffer.c: In function 'psbfb_vm_fault':
>>> drivers/gpu//drm/gma500/framebuffer.c:143:9: warning: 'ret' may be used uninitialized in this function [-Wmaybe-uninitialized]
> return ret;
> ^~~
Sorrry about it. Will fix and send v3.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2018-05-26 13:32 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-05-23 14:46 [PATCH v2] gpu: drm: gma500: Change return type to vm_fault_t Souptick Joarder
2018-05-26 2:41 ` kbuild test robot
2018-05-26 13:32 ` Souptick Joarder
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).