Patches to add MODULE_VERSION lines to several network drivers...
Here is the list:
tg3
e100
e1000
b44
tulip
3c59x
8139too
ns83820
r8169
bonding
vlan
Individual patches to follow...
John
--
John W. Linville
[email protected]
Add MODULE_VERSION to e100 driver.
drivers/net/e100.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/e100.c.orig
+++ linux-2.6.9/drivers/net/e100.c
@@ -166,6 +166,7 @@
MODULE_DESCRIPTION(DRV_DESCRIPTION);
MODULE_AUTHOR(DRV_COPYRIGHT);
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
static int debug = 3;
module_param(debug, int, 0);
Add MODULE_VERSION to e1000 driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/e1000/e1000_main.c | 4 +++-
1 files changed, 3 insertions(+), 1 deletion(-)
Re-send -- forgot Signed-off-by line...
--- linux-2.6.9/drivers/net/e1000/e1000_main.c.orig
+++ linux-2.6.9/drivers/net/e1000/e1000_main.c
@@ -48,7 +48,8 @@ char e1000_driver_string[] = "Intel(R) P
#else
#define DRIVERNAPI "-NAPI"
#endif
-char e1000_driver_version[] = "5.3.19-k2"DRIVERNAPI;
+#define DRV_VERSION "5.3.19-k2"DRIVERNAPI
+char e1000_driver_version[] = DRV_VERSION;
char e1000_copyright[] = "Copyright (c) 1999-2004 Intel Corporation.";
/* e1000_pci_tbl - PCI Device ID Table
@@ -196,6 +197,7 @@ static struct pci_driver e1000_driver =
MODULE_AUTHOR("Intel Corporation, <[email protected]>");
MODULE_DESCRIPTION("Intel(R) PRO/1000 Network Driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
static int debug = NETIF_MSG_DRV | NETIF_MSG_PROBE;
module_param(debug, int, 0);
Add MODULE_VERSION to r8169 driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/r8169.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/r8169.c.orig
+++ linux-2.6.9/drivers/net/r8169.c
@@ -362,6 +362,7 @@ MODULE_PARM(rx_copybreak, "i");
MODULE_PARM(use_dac, "i");
MODULE_PARM_DESC(use_dac, "Enable PCI DAC. Unsafe on 32 bit PCI slot.");
MODULE_LICENSE("GPL");
+MODULE_VERSION(RTL8169_VERSION);
static int rtl8169_open(struct net_device *dev);
static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev);
Add MODULE_VERSION to ns83820 driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/ns83820.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/ns83820.c.ns83820
+++ linux-2.6.9/drivers/net/ns83820.c
@@ -2106,6 +2106,7 @@ static void __exit ns83820_exit(void)
MODULE_AUTHOR("Benjamin LaHaise <[email protected]>");
MODULE_DESCRIPTION("National Semiconductor DP83820 10/100/1000 driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(VERSION);
MODULE_DEVICE_TABLE(pci, ns83820_pci_tbl);
This makes warning error about :
Warning: could not find versions for .tmp_versions/r8169.mod
Do I have to enable something in .config (like CONFIG_MODVERSIONS)?
If so, how does one make this transparent, to get rid of the
warning if CONFIG_MODVERSIONS is not set?
On Wed, 20 Oct 2004, John W. Linville wrote:
> Add MODULE_VERSION to r8169 driver.
>
> Signed-off-by: John W. Linville <[email protected]>
> ---
>
> drivers/net/r8169.c | 1 +
> 1 files changed, 1 insertion(+)
>
> --- linux-2.6.9/drivers/net/r8169.c.orig
> +++ linux-2.6.9/drivers/net/r8169.c
> @@ -362,6 +362,7 @@ MODULE_PARM(rx_copybreak, "i");
> MODULE_PARM(use_dac, "i");
> MODULE_PARM_DESC(use_dac, "Enable PCI DAC. Unsafe on 32 bit PCI slot.");
> MODULE_LICENSE("GPL");
> +MODULE_VERSION(RTL8169_VERSION);
>
> static int rtl8169_open(struct net_device *dev);
> static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev);
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
Cheers,
Dick Johnson
Penguin : Linux version 2.6.9 on an i686 machine (5537.79 GrumpyMips).
98.36% of all statistics are fiction.
On Wed, Oct 20, 2004 at 04:34:45PM -0400, Richard B. Johnson wrote:
>
> This makes warning error about :
>
> Warning: could not find versions for .tmp_versions/r8169.mod
>
> Do I have to enable something in .config (like CONFIG_MODVERSIONS)?
> If so, how does one make this transparent, to get rid of the
> warning if CONFIG_MODVERSIONS is not set?
Odd...I don't get any such warning, with or without
CONFIG_MODVERSIONS...
MODULE_VERSION is used elsewhere -- do you get that warning
from any other modules? Was this from a clean build?
Send me your .config, and I'll look into it.
Thanks,
John
--
John W. Linville
[email protected]
Add MODULE_VERSION to b44 driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/b44.c | 1 +
1 files changed, 1 insertion(+)
Re-send -- forgot Signed-off-by line...
--- linux-2.6.9/drivers/net/b44.c.orig
+++ linux-2.6.9/drivers/net/b44.c
@@ -79,6 +79,7 @@ MODULE_DESCRIPTION("Broadcom 4400 10/100
MODULE_LICENSE("GPL");
MODULE_PARM(b44_debug, "i");
MODULE_PARM_DESC(b44_debug, "B44 bitmapped debugging message enable value");
+MODULE_VERSION(DRV_MODULE_VERSION);
static int b44_debug = -1; /* -1 == use B44_DEF_MSG_ENABLE as value */
Add MODULE_VERSION to bonding driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/bonding/bond_main.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/bonding/bond_main.c.orig
+++ linux-2.6.9/drivers/net/bonding/bond_main.c
@@ -4700,6 +4700,7 @@ static void __exit bonding_exit(void)
module_init(bonding_init);
module_exit(bonding_exit);
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
MODULE_DESCRIPTION(DRV_DESCRIPTION ", v" DRV_VERSION);
MODULE_AUTHOR("Thomas Davis, [email protected] and many others");
MODULE_SUPPORTED_DEVICE("most ethernet devices");
On Wed, 20 Oct 2004 14:33:58 -0400
"John W. Linville" <[email protected]> wrote:
> Add MODULE_VERSION to tg3 driver.
>
> Signed-off-by: John W. Linville <[email protected]>
Applied, thanks John.
On Wed, 20 Oct 2004 14:36:39 -0400
"John W. Linville" <[email protected]> wrote:
> Add MODULE_VERSION to b44 driver.
>
> Signed-off-by: John W. Linville <[email protected]>
Also applied, thanks John.
Add MODULE_VERSION to tg3 driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/tg3.c | 1 +
1 files changed, 1 insertion(+)
Re-send -- forgot Signed-off-by line...
--- linux-2.6.9/drivers/net/tg3.c.orig
+++ linux-2.6.9/drivers/net/tg3.c
@@ -143,6 +143,7 @@ MODULE_DESCRIPTION("Broadcom Tigon3 ethe
MODULE_LICENSE("GPL");
MODULE_PARM(tg3_debug, "i");
MODULE_PARM_DESC(tg3_debug, "Tigon3 bitmapped debugging message enable value");
+MODULE_VERSION(DRV_MODULE_VERSION);
static int tg3_debug = -1; /* -1 == use TG3_DEF_MSG_ENABLE as value */
Add MODULE_VERSION to vlan driver.
Signed-off-by: John W. Linville <[email protected]>
---
net/8021q/vlan.c | 10 ++++++----
1 files changed, 6 insertions(+), 4 deletions(-)
--- linux-2.6.9/net/8021q/vlan.c.orig
+++ linux-2.6.9/net/8021q/vlan.c
@@ -35,6 +35,8 @@
#include "vlan.h"
#include "vlanproc.h"
+#define DRV_VERSION "1.8"
+
/* Global VLAN variables */
/* Our listing of VLAN group(s) */
@@ -42,8 +44,7 @@ struct hlist_head vlan_group_hash[VLAN_G
#define vlan_grp_hashfn(IDX) ((((IDX) >> VLAN_GRP_HASH_SHIFT) ^ (IDX)) & VLAN_GRP_HASH_MASK)
static char vlan_fullname[] = "802.1Q VLAN Support";
-static unsigned int vlan_version = 1;
-static unsigned int vlan_release = 8;
+static char vlan_version[] = DRV_VERSION;
static char vlan_copyright[] = "Ben Greear <[email protected]>";
static char vlan_buggyright[] = "David S. Miller <[email protected]>";
@@ -84,8 +85,8 @@ static int __init vlan_proto_init(void)
{
int err;
- printk(VLAN_INF "%s v%u.%u %s\n",
- vlan_fullname, vlan_version, vlan_release, vlan_copyright);
+ printk(VLAN_INF "%s v%s %s\n",
+ vlan_fullname, vlan_version, vlan_copyright);
printk(VLAN_INF "All bugs added by %s\n",
vlan_buggyright);
@@ -735,3 +736,4 @@ static int vlan_ioctl_handler(void __use
}
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
Add MODULE_VERSION to e100 driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/e100.c | 1 +
1 files changed, 1 insertion(+)
Re-send -- forgot Signed-off-by line...
--- linux-2.6.9/drivers/net/e100.c.orig
+++ linux-2.6.9/drivers/net/e100.c
@@ -166,6 +166,7 @@
MODULE_DESCRIPTION(DRV_DESCRIPTION);
MODULE_AUTHOR(DRV_COPYRIGHT);
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
static int debug = 3;
module_param(debug, int, 0);
Add MODULE_VERSION to 8139too driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/8139too.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/8139too.c.8139too
+++ linux-2.6.9/drivers/net/8139too.c
@@ -598,6 +598,7 @@ struct rtl8139_private {
MODULE_AUTHOR ("Jeff Garzik <[email protected]>");
MODULE_DESCRIPTION ("RealTek RTL-8139 Fast Ethernet driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
MODULE_PARM (multicast_filter_limit, "i");
MODULE_PARM (media, "1-" __MODULE_STRING(MAX_UNITS) "i");
Add MODULE_VERSION to 3c59x driver.
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/3c59x.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/3c59x.c.orig
+++ linux-2.6.9/drivers/net/3c59x.c
@@ -277,6 +277,7 @@ MODULE_AUTHOR("Donald Becker <becker@scy
MODULE_DESCRIPTION("3Com 3c59x/3c9xx ethernet driver "
DRV_VERSION " " DRV_RELDATE);
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
MODULE_PARM(debug, "i");
MODULE_PARM(global_options, "i");
Add MODULE_VERSION to e1000 driver.
drivers/net/e1000/e1000_main.c | 4 +++-
1 files changed, 3 insertions(+), 1 deletion(-)
--- linux-2.6.9/drivers/net/e1000/e1000_main.c.orig
+++ linux-2.6.9/drivers/net/e1000/e1000_main.c
@@ -48,7 +48,8 @@ char e1000_driver_string[] = "Intel(R) P
#else
#define DRIVERNAPI "-NAPI"
#endif
-char e1000_driver_version[] = "5.3.19-k2"DRIVERNAPI;
+#define DRV_VERSION "5.3.19-k2"DRIVERNAPI
+char e1000_driver_version[] = DRV_VERSION;
char e1000_copyright[] = "Copyright (c) 1999-2004 Intel Corporation.";
/* e1000_pci_tbl - PCI Device ID Table
@@ -196,6 +197,7 @@ static struct pci_driver e1000_driver =
MODULE_AUTHOR("Intel Corporation, <[email protected]>");
MODULE_DESCRIPTION("Intel(R) PRO/1000 Network Driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
static int debug = NETIF_MSG_DRV | NETIF_MSG_PROBE;
module_param(debug, int, 0);
Add MODULE_VERSION to tg3 driver.
drivers/net/tg3.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/tg3.c.orig
+++ linux-2.6.9/drivers/net/tg3.c
@@ -143,6 +143,7 @@ MODULE_DESCRIPTION("Broadcom Tigon3 ethe
MODULE_LICENSE("GPL");
MODULE_PARM(tg3_debug, "i");
MODULE_PARM_DESC(tg3_debug, "Tigon3 bitmapped debugging message enable value");
+MODULE_VERSION(DRV_MODULE_VERSION);
static int tg3_debug = -1; /* -1 == use TG3_DEF_MSG_ENABLE as value */
Add MODULE_VERSION to b44 driver.
drivers/net/b44.c | 1 +
1 files changed, 1 insertion(+)
--- linux-2.6.9/drivers/net/b44.c.orig
+++ linux-2.6.9/drivers/net/b44.c
@@ -79,6 +79,7 @@ MODULE_DESCRIPTION("Broadcom 4400 10/100
MODULE_LICENSE("GPL");
MODULE_PARM(b44_debug, "i");
MODULE_PARM_DESC(b44_debug, "B44 bitmapped debugging message enable value");
+MODULE_VERSION(DRV_MODULE_VERSION);
static int b44_debug = -1; /* -1 == use B44_DEF_MSG_ENABLE as value */
On Wed, 2004-10-20 at 20:11, John W. Linville wrote:
> Patches to add MODULE_VERSION lines to several network drivers...
>
> Here is the list:
have you checked if the version of these drivers is actually useful? (eg
updated when the driver changes) If it's not I'd say adding a
MODULE_VERSION to it makes no sense whatsoever.
Add MODULE_VERSION to the tulip-based drivers
Signed-off-by: John W. Linville <[email protected]>
---
drivers/net/tulip/de2104x.c | 1 +
drivers/net/tulip/dmfe.c | 1 +
drivers/net/tulip/tulip_core.c | 1 +
drivers/net/tulip/winbond-840.c | 1 +
drivers/net/tulip/xircom_tulip_cb.c | 1 +
5 files changed, 5 insertions(+)
--- linux-2.6.9/drivers/net/tulip/de2104x.c.orig
+++ linux-2.6.9/drivers/net/tulip/de2104x.c
@@ -56,6 +56,7 @@ KERN_INFO DRV_NAME " PCI Ethernet driver
MODULE_AUTHOR("Jeff Garzik <[email protected]>");
MODULE_DESCRIPTION("Intel/Digital 21040/1 series PCI Ethernet driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
static int debug = -1;
MODULE_PARM (debug, "i");
--- linux-2.6.9/drivers/net/tulip/tulip_core.c.orig
+++ linux-2.6.9/drivers/net/tulip/tulip_core.c
@@ -115,6 +115,7 @@ static int csr0 = 0x00A00000 | 0x4800;
MODULE_AUTHOR("The Linux Kernel Team");
MODULE_DESCRIPTION("Digital 21*4* Tulip ethernet driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
MODULE_PARM(tulip_debug, "i");
MODULE_PARM(max_interrupt_work, "i");
MODULE_PARM(rx_copybreak, "i");
--- linux-2.6.9/drivers/net/tulip/xircom_tulip_cb.c.orig
+++ linux-2.6.9/drivers/net/tulip/xircom_tulip_cb.c
@@ -116,6 +116,7 @@ KERN_INFO " unofficial 2.4.x kernel port
MODULE_AUTHOR("Donald Becker <[email protected]>");
MODULE_DESCRIPTION("Xircom CBE-100 ethernet driver");
MODULE_LICENSE("GPL v2");
+MODULE_VERSION(DRV_VERSION);
MODULE_PARM(debug, "i");
MODULE_PARM(max_interrupt_work, "i");
--- linux-2.6.9/drivers/net/tulip/dmfe.c.orig
+++ linux-2.6.9/drivers/net/tulip/dmfe.c
@@ -1986,6 +1986,7 @@ static struct pci_driver dmfe_driver = {
MODULE_AUTHOR("Sten Wang, [email protected]");
MODULE_DESCRIPTION("Davicom DM910X fast ethernet driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
MODULE_PARM(debug, "i");
MODULE_PARM(mode, "i");
--- linux-2.6.9/drivers/net/tulip/winbond-840.c.orig
+++ linux-2.6.9/drivers/net/tulip/winbond-840.c
@@ -144,6 +144,7 @@ KERN_INFO " http://www.scyld.com/networ
MODULE_AUTHOR("Donald Becker <[email protected]>");
MODULE_DESCRIPTION("Winbond W89c840 Ethernet driver");
MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
MODULE_PARM(max_interrupt_work, "i");
MODULE_PARM(debug, "i");
On Wed, 20 Oct 2004, John W. Linville wrote:
> On Wed, Oct 20, 2004 at 04:34:45PM -0400, Richard B. Johnson wrote:
>>
>> This makes warning error about :
>>
>> Warning: could not find versions for .tmp_versions/r8169.mod
>>
>> Do I have to enable something in .config (like CONFIG_MODVERSIONS)?
>> If so, how does one make this transparent, to get rid of the
>> warning if CONFIG_MODVERSIONS is not set?
>
> Odd...I don't get any such warning, with or without
> CONFIG_MODVERSIONS...
>
> MODULE_VERSION is used elsewhere -- do you get that warning
> from any other modules? Was this from a clean build?
>
> Send me your .config, and I'll look into it.
>
> Thanks,
>
> John
It's a generic problem with compiling modules outside of
the kernel. The attached module, complete with its Makefile
shows the problem.
We really need to fix such problems because, to use the
same kernel for a development effort means that the kernel
sources are on some NFS server, not writable by developers.
They have write permissions only for their own workstations.
Therefore, one can't copy their sources to the kernel tree,
modify .config and then develop their modules.
Cheers,
Dick Johnson
Penguin : Linux version 2.6.9 on an i686 machine (5537.79 GrumpyMips).
98.36% of all statistics are fiction.
On Thu, Oct 21, 2004 at 11:17:49AM +0200, Arjan van de Ven wrote:
> On Wed, 2004-10-20 at 20:11, John W. Linville wrote:
> > Patches to add MODULE_VERSION lines to several network drivers...
> >
> > Here is the list:
>
> have you checked if the version of these drivers is actually useful? (eg
> updated when the driver changes) If it's not I'd say adding a
> MODULE_VERSION to it makes no sense whatsoever.
Why do I feel like I'm being baited...? :-)
I would have to suspect that if a version string exists, that it has at
least some meaning to the primary developers/maintainters. It certainly
is beyond my control to force the maintainers to give meaning to their
version strings.
Is this a political statement against the MODULE_VERSION macro and/or
its purpose? I'm not overly interested in debating that one...
John
--
John W. Linville
[email protected]
On Thu, 2004-10-21 at 14:22, John W. Linville wrote:
>
> > have you checked if the version of these drivers is actually useful? (eg
> > updated when the driver changes) If it's not I'd say adding a
> > MODULE_VERSION to it makes no sense whatsoever.
>
> Why do I feel like I'm being baited...? :-)
>
> I would have to suspect that if a version string exists, that it has at
> least some meaning to the primary developers/maintainters. It certainly
> is beyond my control to force the maintainers to give meaning to their
> version strings.
Since the skeleton driver includes a define for that, I suspect your
assumption is a bit overly optimistic.
> Is this a political statement against the MODULE_VERSION macro and/or
> its purpose? I'm not overly interested in debating that one...
Not really. I have absolutely no problem with a MODULE_VERSION macro
*IF* the version it advertises means something. However if the version
it advertises has no meaning whatsoever (eg the version number never
gets updated) then imo it's better to NOT advertise anything so that
other tools (like dkms) don't make assumptions and decisions based on
nothing-meaning data.
On Thu, 2004-10-21 at 14:55, John W. Linville wrote:
> > Not really. I have absolutely no problem with a MODULE_VERSION macro
> > *IF* the version it advertises means something. However if the version
> > it advertises has no meaning whatsoever (eg the version number never
> > gets updated) then imo it's better to NOT advertise anything so that
> > other tools (like dkms) don't make assumptions and decisions based on
> > nothing-meaning data.
>
> Again, I think it would have to be the maintainer's responsibility
> to make the version numbers meaningful.
absolutely; however you're not the maintainer of all of the ones you
patched... and I still argue that until the number is meaningful
exporting it as meaning something is more harm than good.
On Thu, Oct 21, 2004 at 03:46:11PM +0200, Arjan van de Ven wrote:
> On Thu, 2004-10-21 at 14:22, John W. Linville wrote:
> > I would have to suspect that if a version string exists, that it has at
> > least some meaning to the primary developers/maintainters. It certainly
> Since the skeleton driver includes a define for that, I suspect your
> assumption is a bit overly optimistic.
Perhaps...still, at least the drivers I touched w/ these patches seem to
have version numbers that are at least somewhat meaningful.
> > Is this a political statement against the MODULE_VERSION macro and/or
> > its purpose? I'm not overly interested in debating that one...
>
> Not really. I have absolutely no problem with a MODULE_VERSION macro
> *IF* the version it advertises means something. However if the version
> it advertises has no meaning whatsoever (eg the version number never
> gets updated) then imo it's better to NOT advertise anything so that
> other tools (like dkms) don't make assumptions and decisions based on
> nothing-meaning data.
Again, I think it would have to be the maintainer's responsibility
to make the version numbers meaningful.
John
--
John W. Linville
[email protected]
On Thu, Oct 21, 2004 at 04:11:17PM +0200, Arjan van de Ven wrote:
> On Thu, 2004-10-21 at 14:55, John W. Linville wrote:
> > Again, I think it would have to be the maintainer's responsibility
> > to make the version numbers meaningful.
>
> absolutely; however you're not the maintainer of all of the ones you
> patched... and I still argue that until the number is meaningful
> exporting it as meaning something is more harm than good.
As you say, I'm not the maintainer. I just proposed the patches.
It seems you are not really talking to me. Instead, you are telling the
maintainers that by applying these patches, they should be committing
themselves to making/keeping the version numbers meaningful?
John
--
John W. Linville
[email protected]