Syslinux - Dec 2014 - [PATCH] check-gnu-efi.sh: print the output of build-gnu-efi.sh

The build would be failed and exit if build-gnu-efi.sh is failed, so
we'd better not suppress the output of build-gnu-efi.sh, which is good
for debugging.

Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
---
 efi/check-gnu-efi.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/efi/check-gnu-efi.sh b/efi/check-gnu-efi.sh
index 85305f8..7d99e9a 100755
--- a/efi/check-gnu-efi.sh
+++ b/efi/check-gnu-efi.sh
@@ -25,7 +25,7 @@ if [ ! \( -f "$objdir/include/efi/$ARCH/efibind.h"
-a -f "$objdir/lib/libefi.a"
     # Syslinux disables built-in implicit rules.
     export MAKEFLAGS 
-    ../../efi/build-gnu-efi.sh $ARCH "$objdir" > /dev/null
2>&1
+    ../../efi/build-gnu-efi.sh $ARCH "$objdir"
     if [ $? -ne 0 ]; then
 	printf "Failed to build gnu-efi. "
 	printf "Execute the following command for full details: \n\n"
-- 
1.9.1

Hi

On Sun, Nov 23, 2014 at 11:36 PM, Robert Yang <liezhi.yang at
windriver.com> wrote:
> The build would be failed and exit if build-gnu-efi.sh is failed, so
> we'd better not suppress the output of build-gnu-efi.sh, which is good
> for debugging.
>
> Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
> ---
>  efi/check-gnu-efi.sh | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/efi/check-gnu-efi.sh b/efi/check-gnu-efi.sh
> index 85305f8..7d99e9a 100755
> --- a/efi/check-gnu-efi.sh
> +++ b/efi/check-gnu-efi.sh
> @@ -25,7 +25,7 @@ if [ ! \( -f
"$objdir/include/efi/$ARCH/efibind.h" -a -f
"$objdir/lib/libefi.a"
>      # Syslinux disables built-in implicit rules.
>      export MAKEFLAGS>
> -    ../../efi/build-gnu-efi.sh $ARCH "$objdir" > /dev/null
2>&1
> +    ../../efi/build-gnu-efi.sh $ARCH "$objdir"
>      if [ $? -ne 0 ]; then
>         printf "Failed to build gnu-efi. "
>         printf "Execute the following command for full details:
\n\n"
+1

It is exactly what we do in Arch linux package
https://projects.archlinux.org/svntogit/packages.git/tree/trunk/PKGBUILD?h=packages/syslinux#n50

building 32bit package on x86_64 was failing for me and it took a
while to find why there are no any error message in log.

> The build would be failed and exit if build-gnu-efi.sh is failed, so
> we'd better not suppress the output of build-gnu-efi.sh, which is good
> for debugging.
> 
> Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
> ---
>  efi/check-gnu-efi.sh | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/efi/check-gnu-efi.sh b/efi/check-gnu-efi.sh
> index 85305f8..7d99e9a 100755
> --- a/efi/check-gnu-efi.sh
> +++ b/efi/check-gnu-efi.sh
> @@ -25,7 +25,7 @@ if [ ! \( -f
"$objdir/include/efi/$ARCH/efibind.h" -a -f
"$objdir/lib/libefi.a"
>      # Syslinux disables built-in implicit rules.
>      export MAKEFLAGS>  
> -    ../../efi/build-gnu-efi.sh $ARCH "$objdir" > /dev/null
2>&1
> +    ../../efi/build-gnu-efi.sh $ARCH "$objdir"
>      if [ $? -ne 0 ]; then
>  	printf "Failed to build gnu-efi. "
>  	printf "Execute the following command for full details: \n\n"
> -- 
> 1.9.1
> 
> _______________________________________________
> Syslinux mailing list
> Submissions to Syslinux at zytor.com
> Unsubscribe or set options at:
> http://www.zytor.com/mailman/listinfo/syslinux
> 
Several comments about the gnu-efi submodule in the Syslinux source 
code (as of official upstream Syslinux 6.03 stable)...

_ For the gnu-efi submodule, 'git' is a dependency.

_ The gnu-efi submodule in official Syslinux points to a specific 
commit of gnu-efi. Pointing to (or rebuilding basing on) a different 
(usually newer) commit might result in different (unknown) behaviors 
(either introducing or solving problems).

_ The gnu-efi git repository is currently partially unavailable. Trying 
to retrieve older commits (as the one used in the official release of 
Syslinux 6.03 stable) might succeed, but, at this moment, trying to 
retrieve the official gnu-efi master head will fail.

_ There are some reports of interaction problems (e.g. gcc 4.9.x and/or 
nasm 2.11.06), preventing to successfully build Syslinux.

Regards,
Ady.

On 11/25/2014 10:45 AM, Ady wrote:
>
>> The build would be failed and exit if build-gnu-efi.sh is failed, so
>> we'd better not suppress the output of build-gnu-efi.sh, which is
good
>> for debugging.
>>
>> Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
>> ---
>>   efi/check-gnu-efi.sh | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/efi/check-gnu-efi.sh b/efi/check-gnu-efi.sh
>> index 85305f8..7d99e9a 100755
>> --- a/efi/check-gnu-efi.sh
>> +++ b/efi/check-gnu-efi.sh
>> @@ -25,7 +25,7 @@ if [ ! \( -f
"$objdir/include/efi/$ARCH/efibind.h" -a -f
"$objdir/lib/libefi.a"
>>       # Syslinux disables built-in implicit rules.
>>       export MAKEFLAGS>>
>> -    ../../efi/build-gnu-efi.sh $ARCH "$objdir" >
/dev/null 2>&1
>> +    ../../efi/build-gnu-efi.sh $ARCH "$objdir"
>>       if [ $? -ne 0 ]; then
>>   	printf "Failed to build gnu-efi. "
>>   	printf "Execute the following command for full details:
\n\n"
>> --
>> 1.9.1
>>
>> _______________________________________________
>> Syslinux mailing list
>> Submissions to Syslinux at zytor.com
>> Unsubscribe or set options at:
>> http://www.zytor.com/mailman/listinfo/syslinux
>>
>
> Several comments about the gnu-efi submodule in the Syslinux source
> code (as of official upstream Syslinux 6.03 stable)...
Thanks for the comments, I think that what the patch modifies is the
syslinux repo, not submodule "gnu-efi".

// Robert
>
> _ For the gnu-efi submodule, 'git' is a dependency.
>
> _ The gnu-efi submodule in official Syslinux points to a specific
> commit of gnu-efi. Pointing to (or rebuilding basing on) a different
> (usually newer) commit might result in different (unknown) behaviors
> (either introducing or solving problems).
>
> _ The gnu-efi git repository is currently partially unavailable. Trying
> to retrieve older commits (as the one used in the official release of
> Syslinux 6.03 stable) might succeed, but, at this moment, trying to
> retrieve the official gnu-efi master head will fail.
>
> _ There are some reports of interaction problems (e.g. gcc 4.9.x and/or
> nasm 2.11.06), preventing to successfully build Syslinux.
>
> Regards,
> Ady.
> _______________________________________________
> Syslinux mailing list
> Submissions to Syslinux at zytor.com
> Unsubscribe or set options at:
> http://www.zytor.com/mailman/listinfo/syslinux
>
>

> _ The gnu-efi git repository is currently partially unavailable. Trying 
> to retrieve older commits (as the one used in the official release of 
> Syslinux 6.03 stable) might succeed, but, at this moment, trying to 
> retrieve the official gnu-efi master head will fail.
 
Update: after several days, the gnu-efi master head is now available 
again.

I do not discard the possibility of seeing it disappearing again in the 
future, as this seems to be a misunderstanding / mismanagement of the 
git repo / git commands.

Regards,
Ady.

On Mon, Nov 24, 2014 at 2:36 AM, Robert Yang <liezhi.yang at
windriver.com> wrote:
> The build would be failed and exit if build-gnu-efi.sh is failed, so
> we'd better not suppress the output of build-gnu-efi.sh, which is good
> for debugging.
>
> Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
> ---
>  efi/check-gnu-efi.sh | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/efi/check-gnu-efi.sh b/efi/check-gnu-efi.sh
> index 85305f8..7d99e9a 100755
> --- a/efi/check-gnu-efi.sh
> +++ b/efi/check-gnu-efi.sh
> @@ -25,7 +25,7 @@ if [ ! \( -f
"$objdir/include/efi/$ARCH/efibind.h" -a -f
"$objdir/lib/libefi.a"
>      # Syslinux disables built-in implicit rules.
>      export MAKEFLAGS>
> -    ../../efi/build-gnu-efi.sh $ARCH "$objdir" > /dev/null
2>&1
> +    ../../efi/build-gnu-efi.sh $ARCH "$objdir"
>      if [ $? -ne 0 ]; then
>         printf "Failed to build gnu-efi. "
>         printf "Execute the following command for full details:
\n\n"
Commit 6e832106 which originally created the shell scripts was the
first instance of this silencing.

Aside from adding a few hundred lines of build output (which could be
beneficial), is there any reason not to remove the null redirects?

-- 
-Gene

On Mon, 22 Dec, at 04:51:19PM, Gene Cumm wrote:
> 
> Commit 6e832106 which originally created the shell scripts was the
> first instance of this silencing.
> 
> Aside from adding a few hundred lines of build output (which could be
> beneficial), is there any reason not to remove the null redirects?
There's no downside that I can see. Go for it.

-- 
Matt Fleming, Intel Open Source Technology Center