On Thu, Sep 29, 2016 at 4:14 PM, Stephen Rothwell [off-list ref] wrote:

Hi Segher,

[Adding more cc's]

On Thu, 29 Sep 2016 14:34:42 -0500 Segher Boessenkool [off-list ref] wrote:

quoted

On Thu, Jul 21, 2016 at 07:52:43PM -0400, Paul Gortmaker wrote:

quoted

A new i386-allmodconfig fail showed up relating to VDSO:

I tried to reproduce it locally with x86-64 build host and could
not, so I wonder if it is a missing HOSTCC vs. CC since next
coverage is power host...

  VDSO2C  arch/x86/entry/vdso/vdso-image-32.c
Error: input is not a shared object
make[4]: *** [arch/x86/entry/vdso/vdso-image-32.c] Error 1
make[3]: *** [arch/x86/entry/vdso] Error 2
make[2]: *** [arch/x86/entry] Error 2

This is caused by building on a BE (better-endian) host.  See patch.


Segher


===
From 7f098efa4d184ab1216c6cf8214c44a3abe50a14 Mon Sep 17 00:00:00 2001
Message-Id: [off-list ref]
From: Segher Boessenkool <redacted>
Date: Thu, 29 Sep 2016 11:51:00 +0000
Subject: [PATCH] x86: Fix building on BE

We need to call GET_LE to read hdr->e_type.

Fixes: 57f90c3dfc75 ("x86/vdso: Error out if the vDSO isn't a valid DSO")

quoted

Signed-off-by: Segher Boessenkool <redacted>

57f90c3dfc75 was merged as part of the last merge window, so this is a
fix for Linus' tree.

quoted

---
 arch/x86/entry/vdso/vdso2c.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/entry/vdso/vdso2c.h b/arch/x86/entry/vdso/vdso2c.h
index 4f74119..3dab75f 100644
--- a/arch/x86/entry/vdso/vdso2c.h
+++ b/arch/x86/entry/vdso/vdso2c.h

@@ -22,7 +22,7 @@ static void BITSFUNC(go)(void *raw_addr, size_t raw_len,

      ELF(Phdr) *pt = (ELF(Phdr) *)(raw_addr + GET_LE(&hdr->e_phoff));

-     if (hdr->e_type != ET_DYN)
+     if (GET_LE(&hdr->e_type) != ET_DYN)
              fail("input is not a shared object\n");

Whoops!

Acked-by: Andy Lutomirski <luto@kernel.org>

quoted

      /* Walk the segment table. */
--
1.9.3

--
Cheers,
Stephen Rothwell

-- 
Andy Lutomirski
AMA Capital Management, LLC