Build error: ‘EM_RISCV’ undeclared

Build error: ‘EM_RISCV’ undeclared

[Stefan Wahren]

Hi,

i tried to compile current torvalds tree
("78e709522d2c012cb0daad2e668506637bffb7c2") for arm/multi_v7_defconfig
and get the following build issue:

  UPD     include/generated/utsrelease.h
scripts/sorttable.c: In function ‘do_file’:
scripts/sorttable.c:352:7: error: ‘EM_RISCV’ undeclared (first use in
this function)
  case EM_RISCV:
       ^
scripts/sorttable.c:352:7: note: each undeclared identifier is reported
only once for each function it appears in

I assume this is caused by:

54fed35fd3939398be292e4090b0b1c5ff2238b4 ("riscv: Enable
BUILDTIME_TABLE_SORT")

Best regards
Stefan

Re: Build error: ‘EM_RISCV’ undeclared

[Miles Chen]

On Sun, 2021-09-12 at 12:29 +0200, Stefan Wahren wrote:
> Hi,
> 
> i tried to compile current torvalds tree
> ("78e709522d2c012cb0daad2e668506637bffb7c2") for
> arm/multi_v7_defconfig
> and get the following build issue:
> 
>   UPD     include/generated/utsrelease.h
> scripts/sorttable.c: In function ‘do_file’:
> scripts/sorttable.c:352:7: error: ‘EM_RISCV’ undeclared (first use in
> this function)
>   case EM_RISCV:
>        ^
> scripts/sorttable.c:352:7: note: each undeclared identifier is
> reported
> only once for each function it appears in
> 
> I assume this is caused by:
> 
> 54fed35fd3939398be292e4090b0b1c5ff2238b4 ("riscv: Enable
> BUILDTIME_TABLE_SORT")

I hit this error too. I add the follow conditional EM_RISCV
definition in sorttable.c to build pass (arm64/defconfig).

I will submit a patch out.

diff --git a/scripts/sorttable.c b/scripts/sorttable.c
index f355869c65cd..6ee4fa882919 100644
--- a/scripts/sorttable.c
+++ b/scripts/sorttable.c
@@ -54,6 +54,10 @@
 #define EM_ARCV2       195
 #endif

+#ifndef EM_RISCV
+#define EM_RISCV       243
+#endif
+
> 
> Best regards
> Stefan
> 
> 

Re: Build error: ‘EM_RISCV’ undeclared

[Jon Hunter]

On 13/09/2021 03:49, Miles Chen wrote:
> On Sun, 2021-09-12 at 12:29 +0200, Stefan Wahren wrote:
>> Hi,
>>
>> i tried to compile current torvalds tree
>> ("78e709522d2c012cb0daad2e668506637bffb7c2") for
>> arm/multi_v7_defconfig
>> and get the following build issue:
>>
>>    UPD     include/generated/utsrelease.h
>> scripts/sorttable.c: In function ‘do_file’:
>> scripts/sorttable.c:352:7: error: ‘EM_RISCV’ undeclared (first use in
>> this function)
>>    case EM_RISCV:
>>         ^
>> scripts/sorttable.c:352:7: note: each undeclared identifier is
>> reported
>> only once for each function it appears in
>>
>> I assume this is caused by:
>>
>> 54fed35fd3939398be292e4090b0b1c5ff2238b4 ("riscv: Enable
>> BUILDTIME_TABLE_SORT")
> 
> I hit this error too. I add the follow conditional EM_RISCV
> definition in sorttable.c to build pass (arm64/defconfig).
> 
> I will submit a patch out.
> 
> diff --git a/scripts/sorttable.c b/scripts/sorttable.c
> index f355869c65cd..6ee4fa882919 100644
> --- a/scripts/sorttable.c
> +++ b/scripts/sorttable.c
> @@ -54,6 +54,10 @@
>   #define EM_ARCV2       195
>   #endif
> 
> +#ifndef EM_RISCV
> +#define EM_RISCV       243
> +#endif
> +


This is still broken in the mainline. Any plans to get this fix merged 
there? I believe it has now been broken for nearly 2 weeks.

Thanks!
Jon

-- 
nvpublic

Re: Build error: ‘EM_RISCV’ undeclared

[Jon Hunter]

On 24/09/2021 13:15, miles.chen@mediatek.com wrote:
> From: Jon Hunter <jonathanh@nvidia.com>
> 
> 
>> This is still broken in the mainline. Any plans to get this fix merged 
>> there? I believe it has now been broken for nearly 2 weeks.
> 
> the fix [1] is in linux-next now, thanks.

Yes I know that. However, we need it in the mainline. It has been broken 
for nearly 2 weeks now and all our automation for mainline is not 
running because we cannot build. Please get this fix into mainline as 
soon as possible.

Jon

-- 
nvpublic

Re: Build error: ‘EM_RISCV’ undeclared

[Stefan Wahren]

Hi,

Am 24.09.21 um 14:55 schrieb Jon Hunter:
>
>
> On 24/09/2021 13:15, miles.chen@mediatek.com wrote:
>> From: Jon Hunter <jonathanh@nvidia.com>
>>
>>
>>> This is still broken in the mainline. Any plans to get this fix
>>> merged there? I believe it has now been broken for nearly 2 weeks.
>>
>> the fix [1] is in linux-next now, thanks.
>
> Yes I know that. However, we need it in the mainline. It has been
> broken for nearly 2 weeks now and all our automation for mainline is
> not running because we cannot build. Please get this fix into mainline
> as soon as possible.

as stated in [1], you can try to update you build environment.

[1] -
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20210924&id=fe7118f4742f8c26e532c63bb498283b73e24d7d

>
> Jon
>

Re: Build error: ‘EM_RISCV’ undeclared

[Jon Hunter]

On 24/09/2021 19:05, Stefan Wahren wrote:
> Hi,
> 
> Am 24.09.21 um 14:55 schrieb Jon Hunter:
>>
>>
>> On 24/09/2021 13:15, miles.chen@mediatek.com wrote:
>>> From: Jon Hunter <jonathanh@nvidia.com>
>>>
>>>
>>>> This is still broken in the mainline. Any plans to get this fix
>>>> merged there? I believe it has now been broken for nearly 2 weeks.
>>>
>>> the fix [1] is in linux-next now, thanks.
>>
>> Yes I know that. However, we need it in the mainline. It has been
>> broken for nearly 2 weeks now and all our automation for mainline is
>> not running because we cannot build. Please get this fix into mainline
>> as soon as possible.
> 
> as stated in [1], you can try to update you build environment.

These are farm build machines so not that easy or quick to update. Given 
the change that introduced the breakage is in the mainline, the fix also 
need to be merged into the mainline. Otherwise we need to revert the 
initial change.

Jon

-- 
nvpublic

Re: Build error: ‘EM_RISCV’ undeclared

[Randy Dunlap]

On 9/24/21 12:01 PM, Jon Hunter wrote:
> 
> On 24/09/2021 19:05, Stefan Wahren wrote:
>> Hi,
>>
>> Am 24.09.21 um 14:55 schrieb Jon Hunter:
>>>
>>>
>>> On 24/09/2021 13:15, miles.chen@mediatek.com wrote:
>>>> From: Jon Hunter <jonathanh@nvidia.com>
>>>>
>>>>
>>>>> This is still broken in the mainline. Any plans to get this fix
>>>>> merged there? I believe it has now been broken for nearly 2 weeks.
>>>>
>>>> the fix [1] is in linux-next now, thanks.

so who merged the patch [1] into linux-next?  Andrew?

Seems to me that Andrew should be cc-ed here [added].


>>> Yes I know that. However, we need it in the mainline. It has been
>>> broken for nearly 2 weeks now and all our automation for mainline is
>>> not running because we cannot build. Please get this fix into mainline
>>> as soon as possible.
>>
>> as stated in [1], you can try to update you build environment.
> 
> These are farm build machines so not that easy or quick to update. Given the change that introduced the breakage is in the mainline, the fix also need to be merged into the mainline. Otherwise we need to revert the initial change.

Yes.

-- 
~Randy