2011-12-20 19:19:00

by Luis R. Rodriguez

[permalink] [raw]
Subject: [PATCH] atheros: force endian checks on atheros wireless drivers

From: Luis R. Rodriguez <[email protected]>

Please NACK nasty patches.

Cc: Kalle Valo <[email protected]>
Signed-off-by: Luis R. Rodriguez <[email protected]>
---
drivers/net/wireless/ath/Makefile | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/drivers/net/wireless/ath/Makefile b/drivers/net/wireless/ath/Makefile
index d121469..d716b74 100644
--- a/drivers/net/wireless/ath/Makefile
+++ b/drivers/net/wireless/ath/Makefile
@@ -11,3 +11,4 @@ ath-objs := main.o \
key.o

ath-$(CONFIG_ATH_DEBUG) += debug.o
+ccflags-y += -D__CHECK_ENDIAN__
--
1.7.4.15.g7811d



2011-12-21 15:14:16

by Johannes Berg

[permalink] [raw]
Subject: Re: [PATCH] atheros: force endian checks on atheros wireless drivers

On Wed, 2011-12-21 at 13:10 +0200, Luciano Coelho wrote:

> > ath-$(CONFIG_ATH_DEBUG) += debug.o
> > +ccflags-y += -D__CHECK_ENDIAN__
>
> Doesn't this force everyone who compiles the modules (even
> non-developers) to have sparse installed? Is it really what we want?

No. It just forces everyone who does use sparse to also use it for
endian checks. I do the same in all code I maintain, take a look :-)

johannes


2011-12-21 11:10:44

by Luciano Coelho

[permalink] [raw]
Subject: Re: [PATCH] atheros: force endian checks on atheros wireless drivers

On Tue, 2011-12-20 at 11:18 -0800, Luis R. Rodriguez wrote:
> From: Luis R. Rodriguez <[email protected]>
>
> Please NACK nasty patches.
>
> Cc: Kalle Valo <[email protected]>
> Signed-off-by: Luis R. Rodriguez <[email protected]>
> ---
> drivers/net/wireless/ath/Makefile | 1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/Makefile b/drivers/net/wireless/ath/Makefile
> index d121469..d716b74 100644
> --- a/drivers/net/wireless/ath/Makefile
> +++ b/drivers/net/wireless/ath/Makefile
> @@ -11,3 +11,4 @@ ath-objs := main.o \
> key.o
>
> ath-$(CONFIG_ATH_DEBUG) += debug.o
> +ccflags-y += -D__CHECK_ENDIAN__

Doesn't this force everyone who compiles the modules (even
non-developers) to have sparse installed? Is it really what we want?

If that's the case, shouldn't we do this in a higher level, maybe
enforce sparse check for everything? Of course the number of warnings
would explode, but we could at least do it for all the wireless code in
a first phase?

--
Cheers,
Luca.


2011-12-21 15:26:32

by Luciano Coelho

[permalink] [raw]
Subject: Re: [PATCH] atheros: force endian checks on atheros wireless drivers

On Wed, 2011-12-21 at 16:14 +0100, Johannes Berg wrote:
> On Wed, 2011-12-21 at 13:10 +0200, Luciano Coelho wrote:
>
> > > ath-$(CONFIG_ATH_DEBUG) += debug.o
> > > +ccflags-y += -D__CHECK_ENDIAN__
> >
> > Doesn't this force everyone who compiles the modules (even
> > non-developers) to have sparse installed? Is it really what we want?
>
> No. It just forces everyone who does use sparse to also use it for
> endian checks. I do the same in all code I maintain, take a look :-)

Ah, I get it now. sparse will only run if C=[12] is passed in the make
command.

I'll probably add the same in wl12xx.

It would be nice to have a way to enforce developers to run sparse.
Maybe public humiliation works, but it's not the most efficient way. :)

--
Cheers,
Luca.