diff mbox series

x86/vdso: Conditionally export __vdso_sgx_enter_enclave

Message ID 20221109000306.1407357-1-nathan@kernel.org
State New
Headers
Series x86/vdso: Conditionally export __vdso_sgx_enter_enclave |

Commit Message

Nathan Chancellor Nov. 9, 2022, 12:03 a.m. UTC 
  Recently, ld.lld moved from '--undefined-version' to
'--no-undefined-version' as the default, which breaks building the vDSO
when CONFIG_X86_SGX is not set:

  ld.lld: error: version script assignment of 'LINUX_2.6' to symbol '__vdso_sgx_enter_enclave' failed: symbol not defined

__vdso_sgx_enter_enclave is only included in the vDSO when
CONFIG_X86_SGX is set. Only export it if it will be present in the final
object, which clears up the error.

Link: https://github.com/ClangBuiltLinux/linux/issues/1756
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
---

It would be nice if this could be picked up for an -rc release but I
won't argue otherwise.

Alternatively, we could add '--undefined-version' to the vDSO ldflags
but this does not seem unreasonable to me.

 arch/x86/entry/vdso/vdso.lds.S | 2 ++
 1 file changed, 2 insertions(+)


base-commit: f0c4d9fc9cc9462659728d168387191387e903cc

Comments

Nick Desaulniers Nov. 10, 2022, 8:46 p.m. UTC | #1 
On Tue, Nov 8, 2022 at 4:04 PM Nathan Chancellor <nathan@kernel.org> wrote:
>
> Recently, ld.lld moved from '--undefined-version' to
> '--no-undefined-version' as the default, which breaks building the vDSO
> when CONFIG_X86_SGX is not set:
>
>   ld.lld: error: version script assignment of 'LINUX_2.6' to symbol '__vdso_sgx_enter_enclave' failed: symbol not defined
>
> __vdso_sgx_enter_enclave is only included in the vDSO when
> CONFIG_X86_SGX is set. Only export it if it will be present in the final
> object, which clears up the error.
>
> Link: https://github.com/ClangBuiltLinux/linux/issues/1756
> Signed-off-by: Nathan Chancellor <nathan@kernel.org>
> ---
>
> It would be nice if this could be picked up for an -rc release but I
> won't argue otherwise.

Sounds like the lld change got reverted, so this is less urgent now,
but the change still LGTM regardless. Thanks for the patch!

Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Fixes: 8466436952017 ("x86/vdso: Implement a vDSO for Intel SGX enclave call")


>
> Alternatively, we could add '--undefined-version' to the vDSO ldflags
> but this does not seem unreasonable to me.
>
>  arch/x86/entry/vdso/vdso.lds.S | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/arch/x86/entry/vdso/vdso.lds.S b/arch/x86/entry/vdso/vdso.lds.S
> index 4bf48462fca7..e8c60ae7a7c8 100644
> --- a/arch/x86/entry/vdso/vdso.lds.S
> +++ b/arch/x86/entry/vdso/vdso.lds.S
> @@ -27,7 +27,9 @@ VERSION {
>                 __vdso_time;
>                 clock_getres;
>                 __vdso_clock_getres;
> +#ifdef CONFIG_X86_SGX
>                 __vdso_sgx_enter_enclave;
> +#endif
>         local: *;
>         };
>  }
>
> base-commit: f0c4d9fc9cc9462659728d168387191387e903cc
> --
> 2.38.1
>
>
Nathan Chancellor Dec. 9, 2022, 1:46 p.m. UTC | #2 
Ping? The default switch was remerged a couple of days ago so builds
with tip of tree LLVM are going to be broken again:

https://github.com/llvm/llvm-project/commit/241dbd310599e3c1a0f1b0c9ced14c8b8760539e

Cheers,
Nathan

On Tue, Nov 08, 2022 at 05:03:07PM -0700, Nathan Chancellor wrote:
> Recently, ld.lld moved from '--undefined-version' to
> '--no-undefined-version' as the default, which breaks building the vDSO
> when CONFIG_X86_SGX is not set:
> 
>   ld.lld: error: version script assignment of 'LINUX_2.6' to symbol '__vdso_sgx_enter_enclave' failed: symbol not defined
> 
> __vdso_sgx_enter_enclave is only included in the vDSO when
> CONFIG_X86_SGX is set. Only export it if it will be present in the final
> object, which clears up the error.
> 
> Link: https://github.com/ClangBuiltLinux/linux/issues/1756
> Signed-off-by: Nathan Chancellor <nathan@kernel.org>
> ---
> 
> It would be nice if this could be picked up for an -rc release but I
> won't argue otherwise.
> 
> Alternatively, we could add '--undefined-version' to the vDSO ldflags
> but this does not seem unreasonable to me.
> 
>  arch/x86/entry/vdso/vdso.lds.S | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/arch/x86/entry/vdso/vdso.lds.S b/arch/x86/entry/vdso/vdso.lds.S
> index 4bf48462fca7..e8c60ae7a7c8 100644
> --- a/arch/x86/entry/vdso/vdso.lds.S
> +++ b/arch/x86/entry/vdso/vdso.lds.S
> @@ -27,7 +27,9 @@ VERSION {
>  		__vdso_time;
>  		clock_getres;
>  		__vdso_clock_getres;
> +#ifdef CONFIG_X86_SGX
>  		__vdso_sgx_enter_enclave;
> +#endif
>  	local: *;
>  	};
>  }
> 
> base-commit: f0c4d9fc9cc9462659728d168387191387e903cc
> -- 
> 2.38.1
> 
>
Thomas Gleixner Dec. 9, 2022, 1:54 p.m. UTC | #3 
On Fri, Dec 09 2022 at 06:46, Nathan Chancellor wrote:
> Ping? The default switch was remerged a couple of days ago so builds
> with tip of tree LLVM are going to be broken again:
>
> https://github.com/llvm/llvm-project/commit/241dbd310599e3c1a0f1b0c9ced14c8b8760539e

Sorry. That fell through the cracks. I'll take care of it.
Nathan Chancellor Dec. 9, 2022, 3:55 p.m. UTC | #4 
On Fri, Dec 09, 2022 at 02:54:52PM +0100, Thomas Gleixner wrote:
> On Fri, Dec 09 2022 at 06:46, Nathan Chancellor wrote:
> > Ping? The default switch was remerged a couple of days ago so builds
> > with tip of tree LLVM are going to be broken again:
> >
> > https://github.com/llvm/llvm-project/commit/241dbd310599e3c1a0f1b0c9ced14c8b8760539e
> 
> Sorry. That fell through the cracks. I'll take care of it.

Thanks Thomas! I get you all are always busy so I appreciate the quick
response.

Cheers,
Nathan
diff mbox series

Patch

diff --git a/arch/x86/entry/vdso/vdso.lds.S b/arch/x86/entry/vdso/vdso.lds.S
index 4bf48462fca7..e8c60ae7a7c8 100644
--- a/arch/x86/entry/vdso/vdso.lds.S
+++ b/arch/x86/entry/vdso/vdso.lds.S
@@ -27,7 +27,9 @@  VERSION {
 		__vdso_time;
 		clock_getres;
 		__vdso_clock_getres;
+#ifdef CONFIG_X86_SGX
 		__vdso_sgx_enter_enclave;
+#endif
 	local: *;
 	};
 }