linux-kernel - Re: [PATCH] fbdev driver for S3 Trio/Virge

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Tue, 06 Feb 2007 23:51:06 +0100
From:	Jiri Slaby <jirislaby@...il.com>
To:	Ondrej Zajicek <santiago@...reenet.org>
Cc:	James Simmons <jsimmons@...radead.org>,
	Andrew Morton <akpm@...l.org>, linux-kernel@...r.kernel.org,
	linux-fbdev-devel@...ts.sourceforge.net
Subject: Re: [PATCH] fbdev driver for S3 Trio/Virge

Ondrej Zajicek napsal(a):
> On Mon, Feb 05, 2007 at 11:33:12PM +0100, Jiri Slaby wrote:                                                             
>> Ondrej Zajicek napsal(a):
>>> +#ifdef CONFIG_MTRR
>>> +#include <asm/mtrr.h>
>>> +#endif
>> Why ifdef? It's ifdeffed in the header itself.
> 
> But this header is only available on some archs, isn't it?

Yup, I stand corrected.

>>> +		memset(&(par->state), 0, sizeof(struct vgastate));
>>> +		par->state.flags = VGA_SAVE_MODE | VGA_SAVE_FONTS | 
>>> VGA_SAVE_CMAP;
>>> +		par->state.num_crtc = 0x70;
>>> +		par->state.num_seq = 0x20;
>>> +		save_vga(&(par->state));
>>> +	}
>>> +
>>> +	atomic_inc(&(par->ref_count));
>>> +
>>> +	return 0;
>>> +}
> 
> This pattern is in several fbdev drivers (neofb, i810fb, rivafb, vga16fb).
> I understand the problem but i am not sure what is the best solution.

Blah. It should be fixed, then.

> Maybe acquire/release console semaphore before/after s3fb_open?

Kick atomic_t-s off and use mutexes to protect simple uint or something like this.

>>> +static int s3fb_release(struct fb_info *info, int user)
>>> +{
>>> +	struct s3fb_info *par = (struct s3fb_info *) info;
>> Use container_of for this (and below/above).
> 
> Isn't container_of overkill for this case?

Hmm, maybe yes. Anyway it's not so clean way to do it.

I can see, they have void *par in fb_info. I would do:
struct s3fb_info {
	int chip, rev, mclk_freq;
	int mtrr_reg;
	struct vgastate state;
	atomic_t ref_count;
	u32 pseudo_palette[16];
};
I.e. no struct fb_info inside, since you allocate the space twice by 
framebuffer_alloc() -- 2*sizeof(struct fb_info) + sizeof(struct s3fb_info) + 
padding.
Then normally
framebuffer_alloc(sizeof(struct s3fb_info), NULL);
and happily use
struct s3fb_info *s = info->par;
since par member is set directly to aligned space after fb_info when size is 
nonzero inside the framebuffer_alloc().

Is there any problem with this approach?

regards,
-- 
http://www.fi.muni.cz/~xslaby/            Jiri Slaby
faculty of informatics, masaryk university, brno, cz
e-mail: jirislaby gmail com, gpg pubkey fingerprint:
B674 9967 0407 CE62 ACC8  22A0 32CC 55C3 39D4 7A7E
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/