Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754810Ab2BCIaQ (ORCPT ); Fri, 3 Feb 2012 03:30:16 -0500 Received: from mail-pz0-f46.google.com ([209.85.210.46]:37476 "EHLO mail-pz0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754605Ab2BCIaP (ORCPT ); Fri, 3 Feb 2012 03:30:15 -0500 From: Kautuk Consul To: Russell King Cc: linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, "Mohd. Faris" , Kautuk Consul Subject: [PATCH 1/1] arm: vfp: Raising SIGFPE on invalid floating point operation Date: Fri, 3 Feb 2012 14:06:24 +0530 Message-Id: <1328258184-23082-1-git-send-email-consul.kautuk@gmail.com> X-Mailer: git-send-email 1.7.6 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3612 Lines: 97 There is a lot of ARM/VFP hardware which does not invoke the undefined instruction exception handler (i.e. __und_usr) at all even in case of an invalid floating point operation in usermode. For example, 100.0 divided by 0.0 will not lead to the undefined instruction exception being called by the processor although the hardware does reliably set the DZC bit in the FPSCR. Currently, the usermode code will simply proceed to execute without the SIGFPE signal being raised on that process thus limiting the developer's knowledge of what really went wrong in his code logic. This patch enables the kernel to raise the SIGFPE signal when the faulting thread next calls schedule() in any of the following ways: - When the thread is interrupted by IRQ - When the thread calls a system call - When the thread encounters exception Although the SIGFPE is not delivered at the exact faulting instruction, this patch at least allows the system to handle FPU exception scenarios in a more generic "Linux/Unix" manner. Signed-off-by: Kautuk Consul Signed-off-by: Mohd. Faris --- arch/arm/include/asm/vfp.h | 2 ++ arch/arm/vfp/vfpmodule.c | 15 ++++++++++++++- 2 files changed, 16 insertions(+), 1 deletions(-) diff --git a/arch/arm/include/asm/vfp.h b/arch/arm/include/asm/vfp.h index f4ab34f..a37c265 100644 --- a/arch/arm/include/asm/vfp.h +++ b/arch/arm/include/asm/vfp.h @@ -71,6 +71,8 @@ #define FPSCR_UFC (1<<3) #define FPSCR_IXC (1<<4) #define FPSCR_IDC (1<<7) +#define FPSCR_CUMULATIVE_EXCEPTION_MASK \ + (FPSCR_IOC|FPSCR_DZC|FPSCR_OFC|FPSCR_UFC|FPSCR_IXC|FPSCR_IDC) /* MVFR0 bits */ #define MVFR0_A_SIMD_BIT (0) diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 8f3ccdd..39824a1 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -52,6 +52,8 @@ unsigned int VFP_arch; */ union vfp_state *vfp_current_hw_state[NR_CPUS]; +static void vfp_raise_sigfpe(unsigned int, struct pt_regs *); + /* * Is 'thread's most up to date state stored in this CPUs hardware? * Must be called from non-preemptible context. @@ -72,8 +74,13 @@ static bool vfp_state_in_hw(unsigned int cpu, struct thread_info *thread) */ static void vfp_force_reload(unsigned int cpu, struct thread_info *thread) { + unsigned int fpexc = fmrx(FPEXC); + if (vfp_state_in_hw(cpu, thread)) { - fmxr(FPEXC, fmrx(FPEXC) & ~FPEXC_EN); + if ((fpexc & FPEXC_EN) && + (fmrx(FPSCR) & FPSCR_CUMULATIVE_EXCEPTION_MASK)) + vfp_raise_sigfpe(0, task_pt_regs(current)); + fmxr(FPEXC, fpexc & ~FPEXC_EN); vfp_current_hw_state[cpu] = NULL; } #ifdef CONFIG_SMP @@ -100,6 +107,8 @@ static void vfp_thread_flush(struct thread_info *thread) cpu = get_cpu(); if (vfp_current_hw_state[cpu] == vfp) vfp_current_hw_state[cpu] = NULL; + if (fmrx(FPSCR) & FPSCR_CUMULATIVE_EXCEPTION_MASK) + vfp_raise_sigfpe(0, task_pt_regs(current)); fmxr(FPEXC, fmrx(FPEXC) & ~FPEXC_EN); put_cpu(); @@ -181,6 +190,10 @@ static int vfp_notifier(struct notifier_block *self, unsigned long cmd, void *v) vfp_save_state(vfp_current_hw_state[cpu], fpexc); #endif + if ((fpexc & FPEXC_EN) && + (fmrx(FPSCR) & FPSCR_CUMULATIVE_EXCEPTION_MASK)) + vfp_raise_sigfpe(0, task_pt_regs(current)); + /* * Always disable VFP so we can lazily save/restore the * old state. -- 1.7.6 -- 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/