Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261876AbVCVUbV (ORCPT ); Tue, 22 Mar 2005 15:31:21 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261942AbVCVU3B (ORCPT ); Tue, 22 Mar 2005 15:29:01 -0500 Received: from aun.it.uu.se ([130.238.12.36]:17284 "EHLO aun.it.uu.se") by vger.kernel.org with ESMTP id S261937AbVCVUXh (ORCPT ); Tue, 22 Mar 2005 15:23:37 -0500 Date: Tue, 22 Mar 2005 21:23:30 +0100 (MET) Message-Id: <200503222023.j2MKNUfE011646@harpo.it.uu.se> From: Mikael Pettersson To: linux-kernel@vger.kernel.org Subject: [PATCH][2.4.30-rc1] v10 of gcc4 fixes (solves X crash on x86_64) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3535 Lines: 92 On Sat, 19 Mar 2005 17:26:27 +0100 (MET), I wrote: >Here is a new set of patches to allow gcc-4.0 (20050312) >to compile the 2.4.30-rc1 kernel. ... > The only known problem is >that the X server segfaults during PCI probing on x86-64. I'm >currently debugging that. Problem solved. There was a bug in x86_64's sys_iopl() which allowed the compiler to not perform certain side-effects. The bug was fixed in 2.6.9, as part of a big calling convention change for all system calls that took a struct pt_regs parameter on the stack. v10 of the gcc4 fixes for 2.4.30-rc1 is now available in . An incremental diff for the sys_iopl() fix is included below. /Mikael diff -rupN linux-2.4.30-rc1.gcc4-fixes-v9/arch/x86_64/ia32/ia32entry.S linux-2.4.30-rc1.gcc4-fixes-v10/arch/x86_64/ia32/ia32entry.S --- linux-2.4.30-rc1.gcc4-fixes-v9/arch/x86_64/ia32/ia32entry.S 2005-01-19 18:00:53.000000000 +0100 +++ linux-2.4.30-rc1.gcc4-fixes-v10/arch/x86_64/ia32/ia32entry.S 2005-03-22 21:00:17.000000000 +0100 @@ -113,9 +113,18 @@ quiet_ni_syscall: PTREGSCALL stub32_fork, sys32_fork PTREGSCALL stub32_clone, sys32_clone PTREGSCALL stub32_vfork, sys32_vfork - PTREGSCALL stub32_iopl, sys_iopl PTREGSCALL stub32_rt_sigsuspend, sys_rt_sigsuspend + .macro PTREGSCALL3 label, func, arg + .globl \label +\label: + leaq \func(%rip),%rax + leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */ + jmp ia32_ptregs_common + .endm + + PTREGSCALL3 stub32_iopl, sys_iopl, %rsi + ENTRY(ia32_ptregs_common) popq %r11 SAVE_REST diff -rupN linux-2.4.30-rc1.gcc4-fixes-v9/arch/x86_64/kernel/entry.S linux-2.4.30-rc1.gcc4-fixes-v10/arch/x86_64/kernel/entry.S --- linux-2.4.30-rc1.gcc4-fixes-v9/arch/x86_64/kernel/entry.S 2003-11-29 00:28:11.000000000 +0100 +++ linux-2.4.30-rc1.gcc4-fixes-v10/arch/x86_64/kernel/entry.S 2005-03-22 21:00:17.000000000 +0100 @@ -249,7 +249,16 @@ intret_signal_test: PTREGSCALL stub_vfork, sys_vfork PTREGSCALL stub_rt_sigsuspend, sys_rt_sigsuspend PTREGSCALL stub_sigaltstack, sys_sigaltstack - PTREGSCALL stub_iopl, sys_iopl + + .macro PTREGSCALL3 label,func,arg + .globl \label +\label: + leaq \func(%rip),%rax + leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */ + jmp ptregscall_common + .endm + + PTREGSCALL3 stub_iopl, sys_iopl, %rsi ENTRY(ptregscall_common) popq %r11 diff -rupN linux-2.4.30-rc1.gcc4-fixes-v9/arch/x86_64/kernel/ioport.c linux-2.4.30-rc1.gcc4-fixes-v10/arch/x86_64/kernel/ioport.c --- linux-2.4.30-rc1.gcc4-fixes-v9/arch/x86_64/kernel/ioport.c 2003-11-29 00:28:11.000000000 +0100 +++ linux-2.4.30-rc1.gcc4-fixes-v10/arch/x86_64/kernel/ioport.c 2005-03-22 21:00:17.000000000 +0100 @@ -81,9 +81,9 @@ asmlinkage long sys_ioperm(unsigned long * code. */ -asmlinkage long sys_iopl(unsigned int level, struct pt_regs regs) +asmlinkage long sys_iopl(unsigned int level, struct pt_regs *regs) { - unsigned int old = (regs.eflags >> 12) & 3; + unsigned int old = (regs->eflags >> 12) & 3; if (level > 3) return -EINVAL; @@ -92,6 +92,6 @@ asmlinkage long sys_iopl(unsigned int le if (!capable(CAP_SYS_RAWIO)) return -EPERM; } - regs.eflags = (regs.eflags & 0xffffffffffffcfff) | (level << 12); + regs->eflags = (regs->eflags &~ 0x3000UL) | (level << 12); return 0; } - 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/