On 05/26/2015 10:48 AM, Dan Carpenter wrote:
On Tue, May 26, 2015 at 07:14:01AM -0700, Greg KH wrote:
I take the blame for any problems with Outreachy patches. Given the
huge volume of them, one bug out of 900 isn't that bad of a percentage.
We don't get many bugs through outreachy, but this isn't the first one.
For example, in March and April people complained about:
95745e9b1de2 ('staging: lustre: Use kasprintf.')
45de432775d6 ('Staging: rtl8712: Use memdup_user() instead of copy_from_user()')
There have been others but I have a short memory. We have this
discussion every time. How come no one caught this bug in review?? Oh,
it never went through the list.
I'm fine with 5 bugs per 900 patches or whatever. I wish that the
patches came to the list, but I get that that would double your review
workload to review 900 patches on the outreachy list and again on the
normal dev list.
Adding the maintainer to the original list would still be a good step. I'm not
sure that I would have caught "rtl8712: Use memdup_user() instead of
copy_from_user()" in review, but I certainly would have seen that it was not
proper to do a blanket substitution of del_timer_sync() for a wrapper that used
del_timer().
Obviously the awful coding in rtl8712 contains more traps to trick the novices
than does most code. It also predates a lot of changes in checkpatch.pl and
contains a lot of low-hanging fruit.
Larry
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
