Re: PATCH: PR lto/50568: [4.7 Regression] Massive LTO failures

2011-09-30 Thread Diego Novillo
On Fri, Sep 30, 2011 at 00:09, Andi Kleen wrote: > gcc/lto/: > > 2011-09-29  H.J. Lu   >            Andi Kleen   > >        PR lto/50568 >        * lto.c (lto_splay_tree_delete_id): New. >        (lto_splay_tree_compare_ids): Likewise. >        (lto_splay_tree_lookup): Likewise. >        (lto_spl

Re: PATCH: PR lto/50568: [4.7 Regression] Massive LTO failures

2011-09-29 Thread Andi Kleen
I updated the patch according to your comments. Ok now? -Andi gcc/lto/: 2011-09-29 H.J. Lu Andi Kleen PR lto/50568 * lto.c (lto_splay_tree_delete_id): New. (lto_splay_tree_compare_ids): Likewise. (lto_splay_tree_lookup): Likewise. (lto_s

Re: PATCH: PR lto/50568: [4.7 Regression] Massive LTO failures

2011-09-29 Thread Diego Novillo
On 11-09-29 20:57 , H.J. Lu wrote: diff --git a/gcc/lto/lto.c b/gcc/lto/lto.c index 77eb1a1..9c6770a 100644 --- a/gcc/lto/lto.c +++ b/gcc/lto/lto.c @@ -93,6 +93,93 @@ lto_obj_create_section_hash_table (void) return htab_create (37, hash_name, eq_name, free_with_string); } +/* Delete a al

Re: PATCH: PR lto/50568: [4.7 Regression] Massive LTO failures

2011-09-29 Thread Andi Kleen
"H.J. Lu" writes: > Hi, > > On 32bit hosts, like Linux/ia32, HOST_WIDE_INT is 64bit. But lto and > lto-plugin still uses 32bit int in symbol ID handling. As the result, > LTO is totally broken on Linux/ia32. This patch switches symbol ID > to HOST_WIDE_INT in lto.c and long long in lto-plugin.

PATCH: PR lto/50568: [4.7 Regression] Massive LTO failures

2011-09-29 Thread H.J. Lu
Hi, On 32bit hosts, like Linux/ia32, HOST_WIDE_INT is 64bit. But lto and lto-plugin still uses 32bit int in symbol ID handling. As the result, LTO is totally broken on Linux/ia32. This patch switches symbol ID to HOST_WIDE_INT in lto.c and long long in lto-plugin.c. Since symbol ID is generate