W dniu 10.10.2013 20:05, Michal Nazarewicz pisze:
On Wed, Oct 09 2013, Andrzej Pietrasiewicz wrote:
This will be required by configfs integration.
Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@xxxxxxxxxxx>
Signed-off-by: Kyungin Park <kyungmin.park@xxxxxxxxxxx>
---
drivers/usb/gadget/storage_common.c | 42 +++++++++++++++++++++++++++++++++++
drivers/usb/gadget/storage_common.h | 5 ++++
2 files changed, 47 insertions(+), 0 deletions(-)
diff --git a/drivers/usb/gadget/storage_common.c b/drivers/usb/gadget/storage_common.c
index 969948d..c7b78a1 100644
--- a/drivers/usb/gadget/storage_common.c
+++ b/drivers/usb/gadget/storage_common.c
@@ -439,4 +450,35 @@ ssize_t fsg_store_file(struct fsg_lun *curlun, struct rw_semaphore *filesem,
}
EXPORT_SYMBOL(fsg_store_file);
+ssize_t fsg_store_cdrom(struct fsg_lun *curlun, const char *buf, size_t count)
+{
+ unsigned cdrom;
+ int ret;
+
+ ret = kstrtouint(buf, 2, &cdrom);
Base 2? That's rather strange. Just keep it base 10 or use strtobool().
I vote for strtobool().
+ if (ret)
+ return ret;
+
+ curlun->cdrom = cdrom;
+
+ return count;
+}
+EXPORT_SYMBOL(fsg_store_cdrom);
Loading the module with cdrom set, implies read-only flag, but with this
function, it will become possible to set cdrom without setting read-only
flag. Some care is needed to make it work properly.
Is it enough if I call fsg_store_ro() before actually assigning
curlun->crdrom = cdrom?
Of course some adjustments are needed; the essential part of
fsg_store_ro() factored out to e.g. _fsg_store_ro()
(assumes locks taken), fsg_store_cdrom() accepting the filesem and
taking it before calling _fsg_store_ro(); appropriate error handling in
case _fsg_store_ro() fails.
AP
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
