Hello Rashika, On Fri, Oct 25, 2013 at 7:10 PM, Minchan Kim <minchan@xxxxxxxxxx> wrote: > Hello Rashika, > > First of all, thanks for looking this! > > On Tue, Oct 22, 2013 at 07:00:57PM +0530, Rashika Kheria wrote: >> This patch fixes the following Smatch warning in zram_drv.c- >> drivers/staging/zram/zram_drv.c:663 >> reset_store() warn: variable dereferenced before check 'bdev' (see line 652) >> drivers/staging/zram/zram_drv.c:899 >> destroy_device() warn: variable dereferenced before check 'zram->disk' (see line 896) >> >> Signed-off-by: Rashika Kheria <rashika.kheria@xxxxxxxxx> >> --- >> >> This revision fixes the following issues of the previous revision: >> Remove unnecessary checks >> >> drivers/staging/zram/zram_drv.c | 16 +++++++--------- >> 1 file changed, 7 insertions(+), 9 deletions(-) >> >> diff --git a/drivers/staging/zram/zram_drv.c b/drivers/staging/zram/zram_drv.c >> index 2c4ed52..8679a06 100644 >> --- a/drivers/staging/zram/zram_drv.c >> +++ b/drivers/staging/zram/zram_drv.c >> @@ -648,6 +648,9 @@ static ssize_t reset_store(struct device *dev, >> zram = dev_to_zram(dev); >> bdev = bdget_disk(zram->disk, 0); >> >> + if (!bdev) >> + return -EBUSY; >> + > > I'd like to look into that when bdget_disk could fail but I don't have a time > so sorry. I will review it when I return back to the office in next week. I reviewed that part and realized bdget_disk could fail when memory is tight so you're right. But please separate it with two stuffs. 1. fix reset_store bug caused by accessing NULL ptr 2. fix destroy_device by smatch warning. 1 is real bug fix which NULL ptr access will happen when memory is tight although it's very rare so that IMO, it should go stable tree while 2 is just smatch warning which will never happen in real practice. Thanks for looking this! -- Kind regards, Minchan Kim _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
