[BUG+PATCH] binfmt_elf::create_elf_tables changes u_platform without refill

From: J.A. Magallon (jamagallon@able.es)
Date: Fri Dec 27 2002 - 18:58:35 EST


Hi all...

I thing I found a bug in 2.4 (and someone should check if this applies to 2.5
also...). In short, when running on an HT box, create_elf_tables() changes
u_platform pointer (to separate stacks and avoid cache wars), without
copy_to_user'ing again.

Current code:

    if (k_platform) {
        platform_len = strlen(k_platform) + 1;
        u_platform = p - platform_len;
        __copy_to_user(u_platform, k_platform, platform_len);
    } else
...
#if defined(__i386__) && defined(CONFIG_SMP)
    if(smp_num_siblings > 1)
        u_platform = u_platform - ((current->pid % 64) << 7);
///// So original u_platform with data is lost !!!
#endif
...
    sp = (elf_addr_t *)(~15UL & (unsigned long)(u_platform)); // Set stack
...
    if (k_platform) {
        sp -= 2;
        NEW_AUX_ENT(0, AT_PLATFORM, (elf_addr_t)(unsigned long) u_platform);
///// What the h**l is copied here after changing u_platform ???
    }

Patch below (use an independent stack_top). Boots and works on my dual Xeon,
and glibc behaves correctly:

--- linux/fs/binfmt_elf.c.orig 2002-12-28 00:12:32.000000000 +0100
+++ linux/fs/binfmt_elf.c 2002-12-28 00:32:37.000000000 +0100
@@ -116,11 +116,14 @@
         elf_caddr_t *argv;
         elf_caddr_t *envp;
         elf_addr_t *sp, *csp;
+ char *stack_top;
         char *k_platform, *u_platform;
         long hwcap;
         size_t platform_len = 0;
         size_t len;
 
+ stack_top = p;
+
         /*
          * Get hold of platform and hardware capabilities masks for
          * the machine we are running on. In some cases (Sparc),
@@ -135,8 +138,8 @@
                 platform_len = strlen(k_platform) + 1;
                 u_platform = p - platform_len;
                 __copy_to_user(u_platform, k_platform, platform_len);
- } else
- u_platform = p;
+ stack_top = u_platform;
+ }
 
 #if defined(__i386__) && defined(CONFIG_SMP)
         /*
@@ -149,15 +152,14 @@
          * processors. This keeps Mr Marcelo Person happier but should be
          * removed for 2.5
          */
-
         if(smp_num_siblings > 1)
- u_platform = u_platform - ((current->pid % 64) << 7);
+ stack_top -= ((current->pid % 64) << 7);
 #endif
 
         /*
          * Force 16 byte _final_ alignment here for generality.
          */
- sp = (elf_addr_t *)(~15UL & (unsigned long)(u_platform));
+ sp = (elf_addr_t *)(~15UL & (unsigned long)(stack_top));
         csp = sp;
         csp -= (1+DLINFO_ITEMS)*2 + (k_platform ? 2 : 0);
 #ifdef DLINFO_ARCH_ITEMS

One question: why the 64 value ? Because linux supports 64 processors ?
Why not just 2, because you have two siblings at most ?

TIA

-- 
J.A. Magallon <jamagallon@able.es>      \                 Software is like sex:
werewolf.able.es                         \           It's better when it's free
Mandrake Linux release 9.1 (Cooker) for i586
Linux 2.4.21-pre2-jam1 (gcc 3.2.1 (Mandrake Linux 9.1 3.2.1-2mdk))
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Tue Dec 31 2002 - 22:00:11 EST