[klibc 12/43] Enable CONFIG_KLIBC_ZLIB (now required to build kinit)

[klibc 12/43] Enable CONFIG_KLIBC_ZLIB (now required to build kinit)

[H. Peter Anvin]

After removing the private copy of zlib in kinit, we need
CONFIG_KLIBC_ZLIB in order to build klibc.  zlib is required in order
to decompress classical ramdisks.

In the future this should maybe be made conditional on CONFIG_BLK_DEV_RAM.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>

---
commit baacd5d81ff7151e8df3893850ec363441886a1e
tree 0075da4cf5c328f29a89c98e053a5b2cd5e50afe
parent d669be433cc9bc413d958e140f8ec47211498a2a
author H. Peter Anvin <hpa@zytor.com> Mon, 01 May 2006 02:51:52 -0700
committer H. Peter Anvin <hpa@zytor.com> Sun, 18 Jun 2006 18:47:36 -0700

 usr/Kconfig |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/usr/Kconfig b/usr/Kconfig
index ea3b520..4103a02 100644
--- a/usr/Kconfig
+++ b/usr/Kconfig
@@ -48,3 +48,7 @@ config INITRAMFS_ROOT_GID
 config KLIBC_ERRLIST
 	bool
 	default y
+
+config KLIBC_ZLIB
+	bool
+	default y

Re: [klibc 12/43] Enable CONFIG_KLIBC_ZLIB (now required to build kinit)

[Jeff Garzik]

H. Peter Anvin wrote:
> After removing the private copy of zlib in kinit, we need
> CONFIG_KLIBC_ZLIB in order to build klibc.  zlib is required in order
> to decompress classical ramdisks.
> 
> In the future this should maybe be made conditional on CONFIG_BLK_DEV_RAM.

> +config KLIBC_ZLIB
> +	bool
> +	default y


Dumb question, then:  why not make it conditional on rd now?

	Jeff

Re: [klibc 12/43] Enable CONFIG_KLIBC_ZLIB (now required to build kinit)

[H. Peter Anvin]

Jeff Garzik wrote:
> H. Peter Anvin wrote:
>> After removing the private copy of zlib in kinit, we need
>> CONFIG_KLIBC_ZLIB in order to build klibc.  zlib is required in order
>> to decompress classical ramdisks.
>>
>> In the future this should maybe be made conditional on 
>> CONFIG_BLK_DEV_RAM.
> 
>> +config KLIBC_ZLIB
>> +    bool
>> +    default y
> 
> 
> Dumb question, then:  why not make it conditional on rd now?
> 

Mostly because I haven't yet broken down kinit in configurable chunks; 
it's somewhat messy to do as long as I'm maintaining a standalone 
distribution as well as the out-of-mainline kernel tree.  Not difficult 
by any means, but I've wanted to avoid combinatorics and code churn as 
long as I have a limited testing base (although -mm has vastly increased 
it.)

It's also unclear to me if it's the right thing to omit it from the 
library, ever; the only users that are affected are the ones that are 
using the shared klibc and don't have any use of zlib.

	-hpa