[v2] fpga: dfl: afu: support debug access to memory-mapped afu regions
diff mbox series

Message ID 20200304133847.103881-1-d.c.ddcc@gmail.com
State New
Headers show
Series
  • [v2] fpga: dfl: afu: support debug access to memory-mapped afu regions
Related show

Commit Message

Dominic Chen March 4, 2020, 1:38 p.m. UTC
Allow debug access to memory-mapped regions using e.g. gdb.

Signed-off-by: Dominic Chen <d.c.ddcc@gmail.com>
---
 drivers/fpga/dfl-afu-main.c | 9 +++++++++
 1 file changed, 9 insertions(+)

Comments

Moritz Fischer March 5, 2020, 3:09 a.m. UTC | #1
On Wed, Mar 04, 2020 at 08:38:47AM -0500, Dominic Chen wrote:
> Allow debug access to memory-mapped regions using e.g. gdb.
> 
> Signed-off-by: Dominic Chen <d.c.ddcc@gmail.com>
> ---
>  drivers/fpga/dfl-afu-main.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/drivers/fpga/dfl-afu-main.c b/drivers/fpga/dfl-afu-main.c
> index 02baa6a227c0..ec9dbd3d51b8 100644
> --- a/drivers/fpga/dfl-afu-main.c
> +++ b/drivers/fpga/dfl-afu-main.c
> @@ -459,6 +459,12 @@ static long afu_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
>  	return -EINVAL;
>  }
>  
> +static const struct vm_operations_struct afu_vma_ops = {
> +#ifdef CONFIG_HAVE_IOREMAP_PROT
> +	.access = generic_access_phys,
> +#endif /* CONFIG_HAVE_IOREMAP_PROT */
I remember there were issues in the past with modules and this, are
those resolved?

> +};
> +
>  static int afu_mmap(struct file *filp, struct vm_area_struct *vma)
>  {
>  	struct platform_device *pdev = filp->private_data;
> @@ -488,6 +494,9 @@ static int afu_mmap(struct file *filp, struct vm_area_struct *vma)
>  	    !(region.flags & DFL_PORT_REGION_WRITE))
>  		return -EPERM;
>  
> +	// Support debug access to the mapping

Please use kernel style comments /* */ checkpatch.pl should catch this.
> +	vma->vm_ops = &afu_vma_ops;
> +
>  	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
>  
>  	return remap_pfn_range(vma, vma->vm_start,
> -- 
> 2.17.1
> 

Thanks,
Moritz
Dominic Chen March 5, 2020, 3:32 a.m. UTC | #2
On 3/4/2020 10:09 PM, Moritz Fischer wrote:
>> +static const struct vm_operations_struct afu_vma_ops = {
>> +#ifdef CONFIG_HAVE_IOREMAP_PROT
>> +	.access = generic_access_phys,
>> +#endif /* CONFIG_HAVE_IOREMAP_PROT */
> I remember there were issues in the past with modules and this, are
> those resolved?

Are you referring to the `generic_access_phys` symbol not being
exported? It looks like it became available to GPL modules starting with
kernel v3.12.

>>  static int afu_mmap(struct file *filp, struct vm_area_struct *vma)
>>  {
>>  	struct platform_device *pdev = filp->private_data;
>> @@ -488,6 +494,9 @@ static int afu_mmap(struct file *filp, struct vm_area_struct *vma)
>>  	    !(region.flags & DFL_PORT_REGION_WRITE))
>>  		return -EPERM;
>>  
>> +	// Support debug access to the mapping
> Please use kernel style comments /* */ checkpatch.pl should catch this.
Hmm, I did run checkpatch.pl, though it contains `$allow_c99_comments =
1`.  I'll send out v3 to fix this.

Thanks,

Dominic

Patch
diff mbox series

diff --git a/drivers/fpga/dfl-afu-main.c b/drivers/fpga/dfl-afu-main.c
index 02baa6a227c0..ec9dbd3d51b8 100644
--- a/drivers/fpga/dfl-afu-main.c
+++ b/drivers/fpga/dfl-afu-main.c
@@ -459,6 +459,12 @@  static long afu_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	return -EINVAL;
 }
 
+static const struct vm_operations_struct afu_vma_ops = {
+#ifdef CONFIG_HAVE_IOREMAP_PROT
+	.access = generic_access_phys,
+#endif /* CONFIG_HAVE_IOREMAP_PROT */
+};
+
 static int afu_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct platform_device *pdev = filp->private_data;
@@ -488,6 +494,9 @@  static int afu_mmap(struct file *filp, struct vm_area_struct *vma)
 	    !(region.flags & DFL_PORT_REGION_WRITE))
 		return -EPERM;
 
+	// Support debug access to the mapping
+	vma->vm_ops = &afu_vma_ops;
+
 	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
 
 	return remap_pfn_range(vma, vma->vm_start,