I don't think this is the right thing because it is needless code. Overall it doesn't really simplify anything. You are worried that reviewers will be confused and think there is a leak in et131x_rx_dma_memory_alloc() and then add a kfree() which breaks the code. I think you are right that a reviewer would initially wonder why the code does not call kfree() but then you do a search for fb[0] and it becomes obvious. If we add your patch and the reviewer does a search for fb[0] then it is confusing what the right thing to do is. These kinds of "free later" code, are not the most common way to do things in the kernel but they are used other places as well. regards, dan carpenter _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
