2024-05-29 02:11:54

by John Hubbard

[permalink] [raw]
Subject: [PATCH 0/2] selftests/lib.mk: LLVM=1, CC=clang, and warnings

The kselftests may be built in a couple different ways:
make LLVM=1
make CC=clang

In order to handle both cases, set LLVM=1 if CC=clang. That way,the rest
of lib.mk, and any Makefiles that include lib.mk, can base decisions
solely on whether or not LLVM is set.

Then, build upon that to disable a pair of clang warnings that are
already silenced on gcc.

Doing it this way is much better than the piecemeal approach that I
started with in [1] and [2]. Thanks to Nathan Chancellor for the patch
reviews that led to this approach.

[1] https://lore.kernel.org/[email protected]
[2] https://lore.kernel.org/[email protected]

John Hubbard (2):
selftests/lib.mk: handle both LLVM=1 and CC=clang builds
selftests/lib.mk: silence some clang warnings that gcc already ignores

tools/testing/selftests/lib.mk | 18 ++++++++++++++++++
1 file changed, 18 insertions(+)


base-commit: e0cce98fe279b64f4a7d81b7f5c3a23d80b92fbc
--
2.45.1



2024-05-29 02:12:13

by John Hubbard

[permalink] [raw]
Subject: [PATCH 2/2] selftests/lib.mk: silence some clang warnings that gcc already ignores

gcc defaults to silence (off) for the following warnings, but clang
defaults to the opposite. These warnings are not useful for kselftests,
so silence them for the clang builds as well:

-Wno-address-of-packed-member
-Wno-gnu-variable-sized-type-not-at-end

This eliminates warnings for the net/ and user_events/ kselftest
subsystems, in these files:

./net/af_unix/scm_rights.c
./net/timestamping.c
./net/ipsec.c
./user_events/perf_test.c

Cc: Nathan Chancellor <[email protected]>
Signed-off-by: John Hubbard <[email protected]>
---
tools/testing/selftests/lib.mk | 6 ++++++
1 file changed, 6 insertions(+)

diff --git a/tools/testing/selftests/lib.mk b/tools/testing/selftests/lib.mk
index 2902787b89b2..41e879f3f8a2 100644
--- a/tools/testing/selftests/lib.mk
+++ b/tools/testing/selftests/lib.mk
@@ -50,6 +50,12 @@ else
CLANG_FLAGS += --target=$(notdir $(CROSS_COMPILE:%-=%))
endif # CROSS_COMPILE

+# gcc defaults to silence (off) for the following warnings, but clang defaults
+# to the opposite. These warnings are not useful for kselftests, so silence them
+# for the clang builds as well.
+CFLAGS += -Wno-address-of-packed-member
+CFLAGS += -Wno-gnu-variable-sized-type-not-at-end
+
CC := $(CLANG) $(CLANG_FLAGS) -fintegrated-as
else
CC := $(CROSS_COMPILE)gcc
--
2.45.1


2024-05-29 02:16:06

by John Hubbard

[permalink] [raw]
Subject: [PATCH 1/2] selftests/lib.mk: handle both LLVM=1 and CC=clang builds

The kselftests may be built in a couple different ways:
make LLVM=1
make CC=clang

In order to handle both cases, set LLVM=1 if CC=clang. That way,the rest
of lib.mk, and any Makefiles that include lib.mk, can base decisions
solely on whether or not LLVM is set.

Cc: Nathan Chancellor <[email protected]>
Cc: Ryan Roberts <[email protected]>

Signed-off-by: John Hubbard <[email protected]>
---
tools/testing/selftests/lib.mk | 12 ++++++++++++
1 file changed, 12 insertions(+)

diff --git a/tools/testing/selftests/lib.mk b/tools/testing/selftests/lib.mk
index 429535816dbd..2902787b89b2 100644
--- a/tools/testing/selftests/lib.mk
+++ b/tools/testing/selftests/lib.mk
@@ -1,5 +1,17 @@
# This mimics the top-level Makefile. We do it explicitly here so that this
# Makefile can operate with or without the kbuild infrastructure.
+
+# The kselftests may be built in a couple different ways:
+# make LLVM=1
+# make CC=clang
+#
+# In order to handle both cases, set LLVM=1 if CC=clang. That way,the rest of
+# lib.mk, and any Makefiles that include lib.mk, can base decisions solely on
+# whether or not LLVM is set.
+ifeq ($(CC),clang)
+ LLVM := 1
+endif
+
ifneq ($(LLVM),)
ifneq ($(filter %/,$(LLVM)),)
LLVM_PREFIX := $(LLVM)
--
2.45.1


2024-05-30 05:05:34

by Muhammad Usama Anjum

[permalink] [raw]
Subject: Re: [PATCH 1/2] selftests/lib.mk: handle both LLVM=1 and CC=clang builds

On 5/29/24 7:08 AM, John Hubbard wrote:
> The kselftests may be built in a couple different ways:
> make LLVM=1
> make CC=clang
>
> In order to handle both cases, set LLVM=1 if CC=clang. That way,the rest
> of lib.mk, and any Makefiles that include lib.mk, can base decisions
> solely on whether or not LLVM is set.
>
> Cc: Nathan Chancellor <[email protected]>
> Cc: Ryan Roberts <[email protected]>
>
> Signed-off-by: John Hubbard <[email protected]>
LGTM

Reviewed-by: Muhammad Usama Anjum <[email protected]>

> ---
> tools/testing/selftests/lib.mk | 12 ++++++++++++
> 1 file changed, 12 insertions(+)
>
> diff --git a/tools/testing/selftests/lib.mk b/tools/testing/selftests/lib.mk
> index 429535816dbd..2902787b89b2 100644
> --- a/tools/testing/selftests/lib.mk
> +++ b/tools/testing/selftests/lib.mk
> @@ -1,5 +1,17 @@
> # This mimics the top-level Makefile. We do it explicitly here so that this
> # Makefile can operate with or without the kbuild infrastructure.
> +
> +# The kselftests may be built in a couple different ways:
> +# make LLVM=1
> +# make CC=clang
> +#
> +# In order to handle both cases, set LLVM=1 if CC=clang. That way,the rest of
> +# lib.mk, and any Makefiles that include lib.mk, can base decisions solely on
> +# whether or not LLVM is set.
> +ifeq ($(CC),clang)
> + LLVM := 1
> +endif
> +
> ifneq ($(LLVM),)
> ifneq ($(filter %/,$(LLVM)),)
> LLVM_PREFIX := $(LLVM)

--
BR,
Muhammad Usama Anjum

2024-05-30 14:26:14

by Shuah Khan

[permalink] [raw]
Subject: Re: [PATCH 2/2] selftests/lib.mk: silence some clang warnings that gcc already ignores

On 5/28/24 20:08, John Hubbard wrote:
> gcc defaults to silence (off) for the following warnings, but clang
> defaults to the opposite. These warnings are not useful for kselftests,
> so silence them for the clang builds as well:

Please you add more information on why they aren't useful
for kselftests.

>
> -Wno-address-of-packed-member
> -Wno-gnu-variable-sized-type-not-at-end
>
> This eliminates warnings for the net/ and user_events/ kselftest
> subsystems, in these files:
>
> ./net/af_unix/scm_rights.c
> ./net/timestamping.c
> ./net/ipsec.c
> ./user_events/perf_test.c
>
> Cc: Nathan Chancellor <[email protected]>
> Signed-off-by: John Hubbard <[email protected]>
> ---
> tools/testing/selftests/lib.mk | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/tools/testing/selftests/lib.mk b/tools/testing/selftests/lib.mk
> index 2902787b89b2..41e879f3f8a2 100644
> --- a/tools/testing/selftests/lib.mk
> +++ b/tools/testing/selftests/lib.mk
> @@ -50,6 +50,12 @@ else
> CLANG_FLAGS += --target=$(notdir $(CROSS_COMPILE:%-=%))
> endif # CROSS_COMPILE
>
> +# gcc defaults to silence (off) for the following warnings, but clang defaults
> +# to the opposite. These warnings are not useful for kselftests, so silence them
> +# for the clang builds as well.
> +CFLAGS += -Wno-address-of-packed-member
> +CFLAGS += -Wno-gnu-variable-sized-type-not-at-end
> +
> CC := $(CLANG) $(CLANG_FLAGS) -fintegrated-as
> else
> CC := $(CROSS_COMPILE)gcc

thanks,
-- Shuah


2024-05-30 19:29:21

by John Hubbard

[permalink] [raw]
Subject: Re: [PATCH 2/2] selftests/lib.mk: silence some clang warnings that gcc already ignores

On 5/30/24 7:25 AM, Shuah Khan wrote:
> On 5/28/24 20:08, John Hubbard wrote:
>> gcc defaults to silence (off) for the following warnings, but clang
>> defaults to the opposite. These warnings are not useful for kselftests,
>> so silence them for the clang builds as well:
>
> Please you add more information on why they aren't useful
> for kselftests.

Ah OK. My wording is a little misleading. The warnings are not useful
for the *kernel*, as previous decided by the gcc settings when building
the kernel. And it is only only due to including kernel data structures
in the selftests, that we get the warnings on clang.

So it is not something unique to the selftests. There is nothing that
the selftests' code does that triggers these warnings, other than the
act of including the kernel's data structures.

I can post a v2 to update both the comment and the commit description.


thanks,
--
John Hubbard
NVIDIA

>
>>
>>      -Wno-address-of-packed-member
>>      -Wno-gnu-variable-sized-type-not-at-end
>>
>> This eliminates warnings for the net/ and user_events/ kselftest
>> subsystems, in these files:
>>
>>      ./net/af_unix/scm_rights.c
>>      ./net/timestamping.c
>>      ./net/ipsec.c
>>      ./user_events/perf_test.c
>>
>> Cc: Nathan Chancellor <[email protected]>
>> Signed-off-by: John Hubbard <[email protected]>
>> ---
>>   tools/testing/selftests/lib.mk | 6 ++++++
>>   1 file changed, 6 insertions(+)
>>
>> diff --git a/tools/testing/selftests/lib.mk
>> b/tools/testing/selftests/lib.mk
>> index 2902787b89b2..41e879f3f8a2 100644
>> --- a/tools/testing/selftests/lib.mk
>> +++ b/tools/testing/selftests/lib.mk
>> @@ -50,6 +50,12 @@ else
>>   CLANG_FLAGS     += --target=$(notdir $(CROSS_COMPILE:%-=%))
>>   endif # CROSS_COMPILE
>> +# gcc defaults to silence (off) for the following warnings, but clang
>> defaults
>> +# to the opposite. These warnings are not useful for kselftests, so
>> silence them
>> +# for the clang builds as well.
>> +CFLAGS += -Wno-address-of-packed-member
>> +CFLAGS += -Wno-gnu-variable-sized-type-not-at-end
>> +
>>   CC := $(CLANG) $(CLANG_FLAGS) -fintegrated-as
>>   else
>>   CC := $(CROSS_COMPILE)gcc
>
> thanks,
> -- Shuah
>



2024-05-31 14:30:45

by Shuah Khan

[permalink] [raw]
Subject: Re: [PATCH 2/2] selftests/lib.mk: silence some clang warnings that gcc already ignores

On 5/30/24 13:28, John Hubbard wrote:
> On 5/30/24 7:25 AM, Shuah Khan wrote:
>> On 5/28/24 20:08, John Hubbard wrote:
>>> gcc defaults to silence (off) for the following warnings, but clang
>>> defaults to the opposite. These warnings are not useful for kselftests,
>>> so silence them for the clang builds as well:
>>
>> Please you add more information on why they aren't useful
>> for kselftests.
>
> Ah OK. My wording is a little misleading. The warnings are not useful
> for the *kernel*, as previous decided by the gcc settings when building
> the kernel. And it is only only due to including kernel data structures
> in the selftests, that we get the warnings on clang.
>
> So it is not something unique to the selftests. There is nothing that
> the selftests' code does that triggers these warnings, other than the
> act of including the kernel's data structures.
>
> I can post a v2 to update both the comment and the commit description.
>

Yes please.

thanks,
-- Shuah