Bug#336167: gcc-4.0: breaks kernel builds in random ways.
tags 336167 +patch thanks Sven Luther wrote: [snip] The appended patch reverts a single line of the diff between 4.0.2-2 and 4.0.2-3 and lets the testcase succeed. I don't know that part of gcc enough to judge if it is a valid fix for the problem. Sven, could you test if this fixes also the problem you see? Ok, i will, altough i would need to rebuild gcc, so it will not be before tomorrow that i can test it. The appended patch fixes the problems for mips-linux, tested with a kernel build and builds of ncpfs and mypasswordsafe. Sven, can you test this patch? Thiemo 2005-11-02 Eric Botcazou [EMAIL PROTECTED] PR rtl-optimization/23585 * rtlanal.c (rtx_addr_can_trap_p_1) PLUS: Return 0 for an address that can't trap plus a constant integer, if the mode has zero size. Index: rtlanal.c === RCS file: /cvs/gcc/gcc/gcc/rtlanal.c,v retrieving revision 1.221 diff -u -p -r1.221 rtlanal.c --- rtlanal.c 20 Oct 2005 12:14:29 - 1.221 +++ rtlanal.c 2 Nov 2005 09:32:05 - @@ -267,7 +267,9 @@ rtx_addr_can_trap_p_1 (rtx x, enum machi { HOST_WIDE_INT offset; - if (!STRICT_ALIGNMENT || !unaligned_mems) + if (!STRICT_ALIGNMENT + || !unaligned_mems + || GET_MODE_SIZE (mode) == 0) return 0; offset = INTVAL (XEXP (x, 1)); -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
On Thu, Nov 03, 2005 at 01:28:57PM +0100, Thiemo Seufer wrote: tags 336167 +patch thanks Sven Luther wrote: [snip] The appended patch reverts a single line of the diff between 4.0.2-2 and 4.0.2-3 and lets the testcase succeed. I don't know that part of gcc enough to judge if it is a valid fix for the problem. Sven, could you test if this fixes also the problem you see? Ok, i will, altough i would need to rebuild gcc, so it will not be before tomorrow that i can test it. The appended patch fixes the problems for mips-linux, tested with a kernel build and builds of ncpfs and mypasswordsafe. Sven, can you test this patch? Do i need to apply the above mentioned patch too or just this one ? Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
Sven Luther wrote: On Thu, Nov 03, 2005 at 01:28:57PM +0100, Thiemo Seufer wrote: tags 336167 +patch thanks Sven Luther wrote: [snip] The appended patch reverts a single line of the diff between 4.0.2-2 and 4.0.2-3 and lets the testcase succeed. I don't know that part of gcc enough to judge if it is a valid fix for the problem. Sven, could you test if this fixes also the problem you see? Ok, i will, altough i would need to rebuild gcc, so it will not be before tomorrow that i can test it. The appended patch fixes the problems for mips-linux, tested with a kernel build and builds of ncpfs and mypasswordsafe. Sven, can you test this patch? Do i need to apply the above mentioned patch too or just this one ? Only this one (it is already in upstream, btw.), the previous one was wrong. Thiemo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
On Thu, Nov 03, 2005 at 02:01:18PM +0100, Thiemo Seufer wrote: Sven Luther wrote: On Thu, Nov 03, 2005 at 01:28:57PM +0100, Thiemo Seufer wrote: tags 336167 +patch thanks Sven Luther wrote: [snip] The appended patch reverts a single line of the diff between 4.0.2-2 and 4.0.2-3 and lets the testcase succeed. I don't know that part of gcc enough to judge if it is a valid fix for the problem. Sven, could you test if this fixes also the problem you see? Ok, i will, altough i would need to rebuild gcc, so it will not be before tomorrow that i can test it. The appended patch fixes the problems for mips-linux, tested with a kernel build and builds of ncpfs and mypasswordsafe. Sven, can you test this patch? Do i need to apply the above mentioned patch too or just this one ? Only this one (it is already in upstream, btw.), the previous one was wrong. Ok, will do, but don't expect result soon, i think my box uses days to run all the gcc test cases (failing over all the ppc64 ones obviously). Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
Thiemo Seufer wrote: Sven Luther wrote: Package: gcc-4.0 Version: 4.0.2-3 Severity: grave Justification: renders package unusable Well, i confirm that this problem is also present on powerpc, using gcc-4.0 4.0.2-3 makes the kernel build fail, while using -2 seems to be ok. I have heard people mentioning two other arches where this is the case (m68k and mips i think) on irc (on #debian-release i think even, not sure), but no bug has been filed so i do it now. My powerpc builds failed with : For mips 2.6.12, which built fine with gcc 4.0.2-2: CC [M] fs/reiserfs/tail_conversion.o fs/reiserfs/tail_conversion.c: In function 'direct2indirect': fs/reiserfs/tail_conversion.c:138: internal compiler error: Floating point exception Please submit a full bug report, with preprocessed source if appropriate. See URL:http://gcc.gnu.org/bugs.html for instructions. For Debian GNU/Linux specific bug reporting instructions, see URL:file:///usr/share/doc/gcc-4.0/README.Bugs. make[5]: *** [fs/reiserfs/tail_conversion.o] Error 1 make[4]: *** [fs/reiserfs] Error 2 Sorry, no testcase yet. The attached testcase triggers this bug on mips-linux when compiled with gcc -O2 -mabi=64 -c The appended patch reverts a single line of the diff between 4.0.2-2 and 4.0.2-3 and lets the testcase succeed. I don't know that part of gcc enough to judge if it is a valid fix for the problem. Sven, could you test if this fixes also the problem you see? Thiemo /* gcc -O2 -mabi=64 -c -o foo.o foo.c */ struct offset_v1 { int k_uniqueness; }; struct offset_v2 { long v; } __attribute__ ((__packed__)); struct reiserfs_key { int k_objectid; union { struct offset_v1 k_offset_v1; struct offset_v2 k_offset_v2; } u; }; struct item_head { struct reiserfs_key ih_key; int ih_version; }; static void set_offset_v2_k_type(struct offset_v2 *v2) { v2-v = 1; } static void set_le_key_k_type (int version, struct reiserfs_key * key) { version ? (key-u.k_offset_v1.k_uniqueness = 1) : set_offset_v2_k_type((key-u.k_offset_v2)); } static void set_le_ih_k_type (struct item_head * ih) { set_le_key_k_type((__builtin_constant_p((ih)-ih_version) ? (ih)-ih_version : (ih)-ih_version), (ih-ih_key)); } void boo(struct item_head *ih, const char *body); void direct2indirect(void) { struct item_head *p_le_ih; struct item_head ind_ih; unsigned int unfm_ptr; if (__builtin_expect(32, 0)) __asm__ (break); set_le_ih_k_type (ind_ih); if (__builtin_constant_p(p_le_ih) ? 1 : 2) { (__builtin_constant_p(__builtin_constant_p(1) == 1)); boo(ind_ih, (char *)unfm_ptr); } } --- gcc-4.0-4.0.2.new/src/gcc/reorg.c 2005-09-03 21:09:51.0 +0200 +++ gcc-4.0-4.0.2/src/gcc/reorg.c 2005-10-31 23:16:47.0 +0100 @@ -2653,7 +2653,7 @@ fill_slots_from_thread (rtx insn, rtx co if (!must_annul (condition == const_true_rtx || (! insn_sets_resource_p (trial, opposite_needed, 1) - ! may_trap_or_fault_p (pat + ! may_trap_p (pat { old_trial = trial; trial = try_split (pat, trial, 0);
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
On Tue, Nov 01, 2005 at 01:33:35PM +0100, Thiemo Seufer wrote: Thiemo Seufer wrote: Sven Luther wrote: Package: gcc-4.0 Version: 4.0.2-3 Severity: grave Justification: renders package unusable Well, i confirm that this problem is also present on powerpc, using gcc-4.0 4.0.2-3 makes the kernel build fail, while using -2 seems to be ok. I have heard people mentioning two other arches where this is the case (m68k and mips i think) on irc (on #debian-release i think even, not sure), but no bug has been filed so i do it now. My powerpc builds failed with : For mips 2.6.12, which built fine with gcc 4.0.2-2: CC [M] fs/reiserfs/tail_conversion.o fs/reiserfs/tail_conversion.c: In function 'direct2indirect': fs/reiserfs/tail_conversion.c:138: internal compiler error: Floating point exception Please submit a full bug report, with preprocessed source if appropriate. See URL:http://gcc.gnu.org/bugs.html for instructions. For Debian GNU/Linux specific bug reporting instructions, see URL:file:///usr/share/doc/gcc-4.0/README.Bugs. make[5]: *** [fs/reiserfs/tail_conversion.o] Error 1 make[4]: *** [fs/reiserfs] Error 2 Sorry, no testcase yet. The attached testcase triggers this bug on mips-linux when compiled with gcc -O2 -mabi=64 -c The appended patch reverts a single line of the diff between 4.0.2-2 and 4.0.2-3 and lets the testcase succeed. I don't know that part of gcc enough to judge if it is a valid fix for the problem. Sven, could you test if this fixes also the problem you see? Ok, i will, altough i would need to rebuild gcc, so it will not be before tomorrow that i can test it. Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
Thiemo Seufer wrote: Sven Luther wrote: Package: gcc-4.0 Version: 4.0.2-3 Severity: grave Justification: renders package unusable Well, i confirm that this problem is also present on powerpc, using gcc-4.0 4.0.2-3 makes the kernel build fail, while using -2 seems to be ok. I have heard people mentioning two other arches where this is the case (m68k and mips i think) on irc (on #debian-release i think even, not sure), but no bug has been filed so i do it now. My powerpc builds failed with : For mips 2.6.12, which built fine with gcc 4.0.2-2: CC [M] fs/reiserfs/tail_conversion.o fs/reiserfs/tail_conversion.c: In function 'direct2indirect': fs/reiserfs/tail_conversion.c:138: internal compiler error: Floating point exception Please submit a full bug report, with preprocessed source if appropriate. See URL:http://gcc.gnu.org/bugs.html for instructions. For Debian GNU/Linux specific bug reporting instructions, see URL:file:///usr/share/doc/gcc-4.0/README.Bugs. make[5]: *** [fs/reiserfs/tail_conversion.o] Error 1 make[4]: *** [fs/reiserfs] Error 2 Sorry, no testcase yet. It affects also userland, ncpfs FTBFS for mips/mipsel. Thiemo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
Thiemo Seufer a écrit : Thiemo Seufer wrote: Sven Luther wrote: Package: gcc-4.0 Version: 4.0.2-3 Severity: grave Justification: renders package unusable It affects also userland, ncpfs FTBFS for mips/mipsel. and also sane-backends on mipsel. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
Package: gcc-4.0 Version: 4.0.2-3 Severity: grave Justification: renders package unusable Well, i confirm that this problem is also present on powerpc, using gcc-4.0 4.0.2-3 makes the kernel build fail, while using -2 seems to be ok. I have heard people mentioning two other arches where this is the case (m68k and mips i think) on irc (on #debian-release i think even, not sure), but no bug has been filed so i do it now. My powerpc builds failed with : 08:22 svenl kernel/spinlock.c:72:61: error: macro _spin_lock_irqsave requires 2 arguments, but only 1 given 08:22 svenl kernel/spinlock.c:99:59: error: macro _read_lock_irqsave requires 2 arguments, but only 1 given 08:22 svenl kernel/spinlock.c:126:60: error: macro _write_lock_irqsave requires 2 arguments, but only 1 given 08:22 svenl /bin/sh: line 1: 7269 Done(1) gcc -m32 -E -Wall -Wundef -Wstrict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common -ffreestanding -O2 -fomit-frame-pointer -Iarch/ppc -msoft-float -pipe -ffixed-r2 -mmultiple -mstring -Wa,-maltivec -Wdeclaration-after-statement -Wno-pointer-sign -D__GENKSYMS__ -Wp,-MD,kernel/.spinlock.o.d -nostdinc -isystem /usr/lib/gcc/powerpc-linux-gnu/4.0.3/include -D__KERNEL__ -Iinclude -Iarch/ppc -Iarch/ppc/include -Wall -Wundef -Wstrict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common -ffreestanding -O2 -fomit-frame-pointer -Iarch/ppc -msoft-float -pipe -ffixed-r2 -mmultiple -mstring -Wa,-maltivec -Wdeclaration-after-statement -Wno-pointer-sign -DKBUILD_BASENAME=spinlock -DKBUILD_MODNAME=spinlock kernel/spinlock.c And then later : 08:42 svenl fs/ext2/acl.c:483: error: called object '0u' is not a function 08:42 svenl {standard input}: Assembler messages: 08:42 svenl {standard input}:39: Error: symbol `error' is already defined 08:42 svenl {standard input}:57: Error: symbol `retval' is already defined 08:42 svenl {standard input}:72: Error: symbol `name_index' is already defined 08:42 svenl {standard input}:77: Error: symbol `value' is already defined While a 4.0.2-2 build passed fine. Friendly, Sven Luther -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: powerpc (ppc) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.14-rc5-powerpc Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968) Versions of packages gcc-4.0 depends on: ii binutils 2.16.1cvs20050902-1 The GNU assembler, linker and bina ii cpp-4.0 4.0.2-3 The GNU C preprocessor ii gcc-4.0-base 4.0.2-3 The GNU Compiler Collection (base ii libc62.3.5-7 GNU C Library: Shared libraries an ii libgcc1 1:4.0.2-3 GCC support library Versions of packages gcc-4.0 recommends: ii libc6-dev 2.3.5-7GNU C Library: Development Librari pn libmudflap0-dev none (no description available) -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#336167: gcc-4.0: breaks kernel builds in random ways.
Sven Luther wrote: Package: gcc-4.0 Version: 4.0.2-3 Severity: grave Justification: renders package unusable Well, i confirm that this problem is also present on powerpc, using gcc-4.0 4.0.2-3 makes the kernel build fail, while using -2 seems to be ok. I have heard people mentioning two other arches where this is the case (m68k and mips i think) on irc (on #debian-release i think even, not sure), but no bug has been filed so i do it now. My powerpc builds failed with : For mips 2.6.12, which built fine with gcc 4.0.2-2: CC [M] fs/reiserfs/tail_conversion.o fs/reiserfs/tail_conversion.c: In function 'direct2indirect': fs/reiserfs/tail_conversion.c:138: internal compiler error: Floating point exception Please submit a full bug report, with preprocessed source if appropriate. See URL:http://gcc.gnu.org/bugs.html for instructions. For Debian GNU/Linux specific bug reporting instructions, see URL:file:///usr/share/doc/gcc-4.0/README.Bugs. make[5]: *** [fs/reiserfs/tail_conversion.o] Error 1 make[4]: *** [fs/reiserfs] Error 2 Sorry, no testcase yet. Thiemo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]