[PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Steven Cole]

I got this error for 2.4.6-pre6 for make xconfig

drivers/net/Config.in: 145: can't handle dep_bool/dep_mbool/dep_tristate condition
make[1]: *** [kconfig.tk] Error 1
make[1]: Leaving directory `/usr/src/linux-2.4.6-pre6/scripts'
make: *** [xconfig] Error 2

This may be the proper fix.
Steven

--- linux/drivers/net/Config.in.original	Wed Jun 27 20:39:50 2001
+++ linux/drivers/net/Config.in	Wed Jun 27 20:41:28 2001
@@ -142,7 +142,7 @@
       tristate '  NE/2 (ne2000 MCA version) support' CONFIG_NE2_MCA
       tristate '  IBM LAN Adapter/A support' CONFIG_IBMLANA
    fi
-   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
+   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI $CONFIG_PCI
    if [ "$CONFIG_NET_PCI" = "y" ]; then
       dep_tristate '    AMD PCnet32 PCI support' CONFIG_PCNET32 $CONFIG_PCI
       dep_tristate '    Adaptec Starfire support (EXPERIMENTAL)' CONFIG_ADAPTEC_STARFIRE $CONFIG_PCI $CONFIG_EXPERIMENTAL

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Jeff Garzik]

Steven Cole wrote:
> -   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
> +   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI $CONFIG_PCI

See the "EISA" and "VLB" parts in there?  EISA != PCI

-- 
Jeff Garzik      | Andre the Giant has a posse.
Building 1024    |
MandrakeSoft     |

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Keith Owens]

On Thu, 28 Jun 2001 00:07:13 -0400, 
Jeff Garzik <jgarzik@mandrakesoft.com> wrote:
>Steven Cole wrote:
>> -   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
>> +   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI $CONFIG_PCI
>
>See the "EISA" and "VLB" parts in there?  EISA != PCI

Against 2.4.6-pre6.

Index: 6-pre6.1/drivers/net/Config.in
--- 6-pre6.1/drivers/net/Config.in Thu, 28 Jun 2001 10:34:32 +1000 kaos (linux-2.4/l/c/9_Config.in 1.1.2.2.1.3.1.8 644)
+++ 6-pre6.1(w)/drivers/net/Config.in Thu, 28 Jun 2001 16:07:03 +1000 kaos (linux-2.4/l/c/9_Config.in 1.1.2.2.1.3.1.8 644)
@@ -142,7 +142,11 @@ if [ "$CONFIG_NET_ETHERNET" = "y" ]; the
       tristate '  NE/2 (ne2000 MCA version) support' CONFIG_NE2_MCA
       tristate '  IBM LAN Adapter/A support' CONFIG_IBMLANA
    fi
-   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
+   if [ "$CONFIG_ISA" = "y" -o "$CONFIG_EISA" = "y" -o "$CONFIG_PCI" = "y" ]; then
+     bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
+   else
+     define_bool CONFIG_NET_PCI n
+   fi
    if [ "$CONFIG_NET_PCI" = "y" ]; then
       dep_tristate '    AMD PCnet32 PCI support' CONFIG_PCNET32 $CONFIG_PCI
       dep_tristate '    Adaptec Starfire support (EXPERIMENTAL)' CONFIG_ADAPTEC_STARFIRE $CONFIG_PCI $CONFIG_EXPERIMENTAL

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Andrzej Krzysztofowicz]

Keith Owens wrote:
> On Thu, 28 Jun 2001 00:07:13 -0400, 
> Jeff Garzik <jgarzik@mandrakesoft.com> wrote:
> >Steven Cole wrote:
> >> -   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
> >> +   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI $CONFIG_PCI
> >
> >See the "EISA" and "VLB" parts in there?  EISA != PCI
> 
> Against 2.4.6-pre6.
> 
> Index: 6-pre6.1/drivers/net/Config.in
> --- 6-pre6.1/drivers/net/Config.in Thu, 28 Jun 2001 10:34:32 +1000 kaos (linux-2.4/l/c/9_Config.in 1.1.2.2.1.3.1.8 644)
> +++ 6-pre6.1(w)/drivers/net/Config.in Thu, 28 Jun 2001 16:07:03 +1000 kaos (linux-2.4/l/c/9_Config.in 1.1.2.2.1.3.1.8 644)
> @@ -142,7 +142,11 @@ if [ "$CONFIG_NET_ETHERNET" = "y" ]; the
>        tristate '  NE/2 (ne2000 MCA version) support' CONFIG_NE2_MCA
>        tristate '  IBM LAN Adapter/A support' CONFIG_IBMLANA
>     fi
> -   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
> +   if [ "$CONFIG_ISA" = "y" -o "$CONFIG_EISA" = "y" -o "$CONFIG_PCI" = "y" ]; then

CONFIG_EISA check in this condition is redundant.
Intentionally ?

> +     bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI

Andrzej
-- 
=======================================================================
  Andrzej M. Krzysztofowicz               ankry@mif.pg.gda.pl
  phone (48)(58) 347 14 61
Faculty of Applied Phys. & Math.,   Technical University of Gdansk

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Keith Owens]

On Thu, 28 Jun 2001 10:45:55 +0200 (MET DST), 
Andrzej Krzysztofowicz <ankry@pg.gda.pl> wrote:
>Keith Owens wrote:
>> Index: 6-pre6.1/drivers/net/Config.in
>> -   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
>> +   if [ "$CONFIG_ISA" = "y" -o "$CONFIG_EISA" = "y" -o "$CONFIG_PCI" = "y" ]; then

>CONFIG_EISA check in this condition is redundant.

True, but the line is a cut and paste from higher up in
drivers/net/Config.in.  Even though it is redundant, it is consistent
with the rest of the file.

drivers/net/Config.in needs a major cleanup, lots of the if statements
can go and be replaced by dep_xxx statements, CONFIG_ETHERTAP is marked
obsolete but is tested against experimental, CONFIG_ZNET is marked
experimental but is tested against obsolete, etc.  I started to clean
up the file then decided to leave it, the entire thing is being
replaced in 2.5 anyway.

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Jeff Garzik]

Keith Owens wrote:
> 
> On Thu, 28 Jun 2001 10:45:55 +0200 (MET DST),
> Andrzej Krzysztofowicz <ankry@pg.gda.pl> wrote:
> >Keith Owens wrote:
> >> Index: 6-pre6.1/drivers/net/Config.in
> >> -   dep_bool '  EISA, VLB, PCI and on board controllers' CONFIG_NET_PCI
> >> +   if [ "$CONFIG_ISA" = "y" -o "$CONFIG_EISA" = "y" -o "$CONFIG_PCI" = "y" ]; then
> 
> >CONFIG_EISA check in this condition is redundant.
> 
> True, but the line is a cut and paste from higher up in
> drivers/net/Config.in.  Even though it is redundant, it is consistent
> with the rest of the file.

It is not redundant because in theory CONFIG_EISA could exist without
CONFIG_ISA.

> drivers/net/Config.in needs a major cleanup, lots of the if statements
> can go and be replaced by dep_xxx statements, CONFIG_ETHERTAP is marked
> obsolete but is tested against experimental, CONFIG_ZNET is marked
> experimental but is tested against obsolete, etc.

Why not send me an incremental patch for these cleanups, on top of the
cleanup patch that (I hope!) Andrzej will send, in respond to me last
reply.

2.4 will be around for quite a while, and Config.in cleanups should
continue to go in.  Sure work should be directed towards 2.5..... but if
somebody sends me a 2.4 patch for drivers/net/Config.in cleanup, I sure
as hell will apply it.

-- 
Jeff Garzik      | Andre the Giant has a posse.
Building 1024    |
MandrakeSoft     |

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Keith Owens]

On Thu, 28 Jun 2001 16:40:51 -0400, 
Jeff Garzik <jgarzik@mandrakesoft.com> wrote:
>Keith Owens wrote:
>> True, but the line is a cut and paste from higher up in
>> drivers/net/Config.in.  Even though it is redundant, it is consistent
>> with the rest of the file.
>
>It is not redundant because in theory CONFIG_EISA could exist without
>CONFIG_ISA.

Really?  I checked all the arch config.in files in the kernel and none
allow EISA without ISA first.

Re: [PATCH] 2.4.6-pre6 fix drivers/net/Config.in error

[Maciej W. Rozycki]

On Thu, 28 Jun 2001, Jeff Garzik wrote:

> It is not redundant because in theory CONFIG_EISA could exist without
> CONFIG_ISA.

 Huh?  You can put an ISA card into an EISA slot.

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +