On 20/04/17 18:39, Ian Abbott wrote:
If a process that has mmap'd a COMEDI buffer is being run under a debugger such as GDB, the buffer contents are inaccessible from the debugger. Support the `access()` VM operation to allow the buffer contents to be accessed by another process.Signed-off-by: Ian Abbott <abbo...@mev.co.uk> --- drivers/staging/comedi/comedi_buf.c | 24 ++++++++++++++++++++++++ drivers/staging/comedi/comedi_fops.c | 15 +++++++++++++++ drivers/staging/comedi/comedi_internal.h | 2 ++ 3 files changed, 41 insertions(+) diff --git a/drivers/staging/comedi/comedi_buf.c b/drivers/staging/comedi/comedi_buf.c index 1e1df89b5018..8e9b30b26810 100644 --- a/drivers/staging/comedi/comedi_buf.c +++ b/drivers/staging/comedi/comedi_buf.c @@ -161,6 +161,30 @@ int comedi_buf_map_put(struct comedi_buf_map *bm) return 1; } +/* helper for "access" vm operation */ +int comedi_buf_map_access(struct comedi_buf_map *bm, unsigned long offset, + void *buf, int len, int write) +{ + unsigned int pgoff = offset & ~PAGE_MASK; + unsigned long pg = offset >> PAGE_SHIFT; + int done = 0; + + while (done < len && pg < bm->n_pages) { + int l = min_t(int, len - done, PAGE_SIZE - pgoff); + void *b = bm->page_list[pg].virt_addr + pgoff; + + if (write) + memcpy(b, buf, l); + else + memcpy(buf, b, l); + buf += l; + done += l; + pg++; + pgoff = 0; + } + return done; +} + /* returns s->async->buf_map and increments its kref refcount */ struct comedi_buf_map * comedi_buf_map_from_subdev_get(struct comedi_subdevice *s) diff --git a/drivers/staging/comedi/comedi_fops.c b/drivers/staging/comedi/comedi_fops.c index 92d864fc08ac..2a8a39653c94 100644 --- a/drivers/staging/comedi/comedi_fops.c +++ b/drivers/staging/comedi/comedi_fops.c @@ -2165,9 +2165,24 @@ static void comedi_vm_close(struct vm_area_struct *area) comedi_buf_map_put(bm); } +static int comedi_vm_access(struct vm_area_struct *vma, unsigned long addr, + void *buf, int len, int write) +{ + struct comedi_buf_map *bm = vma->vm_private_data; + unsigned long offset = + addr - vma->vm_start + (vma->vm_pgoff << PAGE_SHIFT); + + if (len < 0) + return -EINVAL; + if (len > vma->vm_end - vma->vm_start) + len = vma->vm_end - vma->vm_start;
Actually, this limit isn't correct as it doesn't take the offset from `vma->vm_start` to `addr` into account. I'll send another version of the patch to sort that out.
-- -=( Ian Abbott @ MEV Ltd. E-mail: <abbo...@mev.co.uk> )=- -=( Web: http://www.mev.co.uk/ )=- _______________________________________________ devel mailing list de...@linuxdriverproject.org http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
