1 |
On Thu, Oct 28, 2004 at 04:16:14PM +0000, Luke-Jr wrote: |
2 |
> > On Thu, Oct 28, 2004 at 02:05:37AM +0000, Luke-Jr wrote: |
3 |
> > > On Thursday 28 October 2004 1:53 am, you wrote: |
4 |
> > > > Are they sysfs design flaws or udev design flaws. The reason I ask, is |
5 |
> > > > that a bunch of kernel developers were talking about rewriting the |
6 |
> > > > current hotplug/udev system. |
7 |
> > |
8 |
> > No we were not. |
9 |
> |
10 |
> I think he meant some Linux kernel developers, not neccesarilly Gentoo, though |
11 |
> I'm not on LKML or any such lists personally. |
12 |
|
13 |
Heh, I meant "we" as in "we kernel developers", myself included... |
14 |
|
15 |
> > > sysfs works great. I use it regardless of udev. The main problem is that |
16 |
> > > udev defeats the entire purpose of modules. Using udev, you must preload |
17 |
> > > any modules you want to use manually. If you do that, you might as well |
18 |
> > > compile them into your kernel. I'd much rather not have the driver for |
19 |
> > > any of my devices loaded until I actually need them. |
20 |
> > |
21 |
> > See the udev FAQ for the answers to this question. In short, no, udev |
22 |
> > does NOT "defeat the entire purpose of modules." Geesh, where do people |
23 |
> > get ideas like this from... |
24 |
> |
25 |
> >From udev FAQ: |
26 |
> >Q: But udev will not automatically load a driver if a /dev node is opened |
27 |
> >when it is not present like devfs will do. |
28 |
> >A: If you really require this functionality, then use devfs. It is still |
29 |
> >present in the kernel." |
30 |
> |
31 |
> In short, the "solution" to that problem according to the udev FAQ is to not |
32 |
> use udev. |
33 |
|
34 |
If you want to rely on such a broken, antiquated system, sure, don't use |
35 |
udev. |
36 |
|
37 |
> >Q: Oh come on, pretty please. It can't be that hard to do. |
38 |
> >A: Such a functionality isn't needed on a properly configured system. All |
39 |
> > devices present on the system should generate hotplug events, loading |
40 |
> > the appropriate driver, and udev will notice and create the |
41 |
> > appropriate device node. If you don't want to keep all drivers for your |
42 |
> > hardware in memory, then use something else to manage your modules |
43 |
> > (scripts, modules.conf, etc.) This is not a task for udev. |
44 |
> |
45 |
> What makes you think I want the drivers loaded just because the device is |
46 |
> connected/available? It may be insignificantly small, but I don't see a |
47 |
> reason to use the RAM neccesary nor decrease the stability of my systems for |
48 |
> something I'm not using. |
49 |
> For example, my motherboard has a parallel port, but I don't want the driver |
50 |
> loaded unless I'm actually using it (which is fairly rare). |
51 |
|
52 |
Great, then have a "load lp module" script that you run to load the |
53 |
driver. Don't rely on accessing the device node to load a module for |
54 |
you. It's just wrong, and is not the way the Linux kernel has been |
55 |
evolving over the past 4 years. See my posts on lkml for more details, |
56 |
I'm not going to go into it again here. |
57 |
|
58 |
> > > > But they would still be using sysfs. Also I am curious as to what |
59 |
> > > > "new" USB Mass Storage driver you are referring to? The current |
60 |
> > > > mainstream kernel's driver was updated about 4 months ago, and still |
61 |
> > > > presents itself as a low level scsi driver to the kernel (meaning it |
62 |
> > > > needs scsi.ko and sd.ko) |
63 |
> > > |
64 |
> > > Well, I noticed they were suddenly "udX" when I moved to 2.6.9... |
65 |
> > |
66 |
> > "ubX", not "udX". And that happened because you selected the block UB |
67 |
> > driver. So you asked the kernel to do this, nothing "sudden" about it |
68 |
> > at all. |
69 |
> |
70 |
> >From kernel configuration: |
71 |
> >Low Performance USB Block driver (BLK_DEV_UB) |
72 |
> > |
73 |
> >This driver supports certain USB attached storage devices |
74 |
> >such as flash keys. |
75 |
> |
76 |
> It was newly available and the description implies that it is a desirable |
77 |
> option. |
78 |
|
79 |
"Low Performance" is a desirable option? "flash keys" for your |
80 |
IDE/SCSI/USB bridge device? Yeah, I admit the documentation isn't the |
81 |
best, but it should work for your device, but be slow. If you don't |
82 |
like it, don't select it to be built. |
83 |
|
84 |
> > > Now that I think of it, I'm not sure I tested 2.6.9 w/o udev, so it |
85 |
> > > might have been a udev naming thing that I just assumed meant the |
86 |
> > > device was a USB disk (w/o emulation)... Either way, it didn't work |
87 |
> > > very well for me (e.g. I couldn't use my IDE->USB adaptor). |
88 |
> > |
89 |
> > Stick with the usb-storage driver if you don't like the ub driver. It's |
90 |
> > not for everyone (as the documentation for the driver states.) |
91 |
> |
92 |
> Usually, stuff that doesn't work is marked EXPERIMENTAL or such. Nothing to |
93 |
> suggest that in the summary. |
94 |
|
95 |
It isn't EXPERIMENTAL. It works just fine for me on a whole lot of USB |
96 |
devices. Pete has done a wonderful job on the driver. And if you have |
97 |
any problems that might be caused by it, please let him know so he can |
98 |
fix it. If he doesn't know, how do you expect it to be fixed. Please |
99 |
file bugs at bugzilla.kernel.org for stuff like this, or use email to |
100 |
the kernel developers. |
101 |
|
102 |
thanks, |
103 |
|
104 |
greg k-h |
105 |
|
106 |
-- |
107 |
gentoo-dev@g.o mailing list |