On Mon, 11 Mar 2002, Greg KH wrote:
> But you still have to get the order correct. USB has to start after
> PCI.
That's true. It will.
> > One further reason is that I would actually like to link drivers/net/irda
> > from within drivers/net, which would move it in front of USB, though, and
> > thus cause the same problem w.r.t USB IrDA dongles.
>
> Ah, yet another good reason to put all of the usb drivers in the
> /drivers/usb directory :)
You do have a point there ;-)
> > If nobody sees a problem with this change, I can come up with the
> > (trivial) patch.
>
> I'd be interested in seeing the patch.
Appended, "bk send" patch attached.
--Kai
# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher.
# This patch includes the following deltas:
# ChangeSet 1.384 -> 1.385
# drivers/usb/usb.c 1.35 -> 1.36
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 02/03/11 [EMAIL PROTECTED] 1.385
# Make USB core init an subsys_initcall, like the other buses.
#
# As it's linked after PCI, usb_init() will be called after pci_init(),
# which is an subsys_initcall, too. Subtle, but right.
# --------------------------------------------
#
diff -Nru a/drivers/usb/usb.c b/drivers/usb/usb.c
--- a/drivers/usb/usb.c Mon Mar 11 19:47:00 2002
+++ b/drivers/usb/usb.c Mon Mar 11 19:47:00 2002
@@ -2614,7 +2614,7 @@
usb_hub_cleanup();
}
-module_init(usb_init);
+subsys_initcall(usb_init);
module_exit(usb_exit);
/*
This BitKeeper patch contains the following changesets:
+
# User: kai
# Host: tp1.ruhr-uni-bochum.de
# Root: /home/kai/kernel/v2.5/linus-2.5.usb
# Patch vers: 1.3
# Patch type: REGULAR
== ChangeSet ==
[EMAIL PROTECTED]|ChangeSet|20020205173056|16047|c1d11a41ed024864
[EMAIL PROTECTED]|ChangeSet|20020308015057|44839
D 1.385 02/03/11 19:46:01-06:00 [EMAIL PROTECTED] +1 -0
B [EMAIL PROTECTED]|ChangeSet|20020205173056|16047|c1d11a41ed024864
C
c Make USB core init an subsys_initcall, like the other buses.
c
c As it's linked after PCI, usb_init() will be called after pci_init(),
c which is an subsys_initcall, too. Subtle, but right.
K 45887
P ChangeSet
------------------------------------------------
0a0
>
>[EMAIL PROTECTED]|drivers/usb/usb.c|20020205174017|45298|b4e3b28195e623ad
>[EMAIL PROTECTED]|drivers/usb/usb.c|20020312014558|22040
== drivers/usb/usb.c ==
[EMAIL PROTECTED]|drivers/usb/usb.c|20020205174017|45298|b4e3b28195e623ad
[EMAIL PROTECTED]|drivers/usb/usb.c|20020228075941|21593
D 1.36 02/03/11 19:45:58-06:00 [EMAIL PROTECTED] +1 -1
B [EMAIL PROTECTED]|ChangeSet|20020205173056|16047|c1d11a41ed024864
C
c Make USB core init an subsys_initcall, like the other buses.
c
c As it's linked after PCI, usb_init() will be called after pci_init(),
c which is an subsys_initcall, too. Subtle, but right.
K 22040
O -rw-rw-r--
P drivers/usb/usb.c
------------------------------------------------
D2617 1
I2617 1
subsys_initcall(usb_init);
# Patch checksum=a0f1ae38
