Messages in this thread | | | From | Stevie Trujillo <> | Subject | Re: [PATCH] ramoops: use module parameters instead of platform data if not available | Date | Sat, 28 May 2011 12:05:00 +0200 |
| |
On Saturday 28 May 2011 11:01:12 you wrote: > From: Marco Stornelli <marco.stornelli@gmail.com> > > Use generic module parameters instead of platform data, if platform > data are not available. This limitation has been introduced with > commit c3b92ce9e75f6353104fc7f8e32fb9fdb2550ad0. > > Signed-off-by: Marco Stornelli <marco.stornelli@gmail.com> > CC: Kyungmin Park <kyungmin.park@samsung.com> > Reported-by: Stevie Trujillo <stevie.trujillo@gmail.com>
Nice work, I think this will fix my problems :) I have some comments - not sure how many of them are sane.
I think the indent is wrong (mixed tabs + spaces) in ramoops_init. Tried to fix it, but my email client just made it worse :p
With this patch, ramoops_platform_data takes precedence over module parameters. Should it maybe be the other way?
I think you can just statically allocate ramoops_platform_data, since it's only 2x(unsigned long)? You will use one more long in .data, but less in .text?
Not related to the patch: Should the printks end with "\n"? If i do printk(KERN_ERR "a"); printk(KERN_ERR "b"); I get two lines, but with printk(KERN_ERR "a"); printk("b"); they end up on the same line. So if another driver did printk without KERN_ after ramoops, they would end up on same line?
-- Stevie Trujillo
| |