From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <5130FA30.5010106@siemens.com> Date: Fri, 01 Mar 2013 19:57:52 +0100 From: Jan Kiszka MIME-Version: 1.0 References: <5130E77E.8030005@siemens.com> In-Reply-To: <5130E77E.8030005@siemens.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: [Xenomai] [PULL] x86/ipipe: Small fixes and patch against BUG on legacy IRQs to APs List-Id: Discussions about the Xenomai project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Philippe Gerum Cc: Xenomai On 2013-03-01 18:38, Jan Kiszka wrote: > The following changes since commit 946ca3977e75316751caecc7182f4e64af6499d5: > > ipipe-core-3.5.7-blackfin-3 (2013-01-22 11:41:34 +0100) > > are available in the git repository at: > git://git.xenomai.org/ipipe-jki.git for-upstream/3.5 > (well, whatever transport works by then) > > The most important fix is the backport of the upstream patch to fix > unhandled legacy PIC IRQs that show up on secondary CPUs. Linux gets > confused about this, I-pipe even raises a BUG in __ipipe_handle_irq. > That's solved now according to user reports. > > The rest deals with non-GPL exports of central symbols, a small fix for > the latency tracer on x86 and cleanups. > > Jan Kiszka (8): > ipipe: Export ipipe_root_only also for non-GPL modules > ipipe: Export __ipipe_spin_unlock_debug also for non-GPL modules > ipipe: Allow local_irq_enable/disable etc. for non-GPL modules > x86/ipipe: Fix irqoff tracing in __ipipe_syscall_root > x86/ipipe: Stop saving/restoring flags in __ipipe_syscall_root > x86/ipipe: Consolidate printing of I-pipe domain in show_regs > ipipe: Convert ipipe_check_irqoff into macro > ipipe: Remove unused get_wall_to_monotonic Sorry, but I got hold of another bug and extended the pull therefor: x86/ipipe: Do not call TRACE_IRQS_OFF before __ipipe_divert_exception arch/x86/kernel/entry_32.S | 4 ++-- arch/x86/kernel/entry_64.S | 8 ++++++-- 2 files changed, 8 insertions(+), 4 deletions(-) Not critical as long as CONFIG_TRACE_IRQFLAGS if off. But when it's on and someone triggers e.g. an INT1 or INT3, we corrupt the root domain state. Jan > > Suresh Siddha (1): > x86, apic: Cleanup cfg->domain setup for legacy interrupts > > arch/x86/kernel/apic/io_apic.c | 20 ++++---------------- > arch/x86/kernel/dumpstack_32.c | 6 +++--- > arch/x86/kernel/dumpstack_64.c | 4 +--- > arch/x86/kernel/ipipe.c | 11 +++++++---- > include/linux/ipipe_debug.h | 10 +++++----- > include/linux/time.h | 1 - > kernel/ipipe/core.c | 10 +++++----- > kernel/time/timekeeping.c | 6 ------ > 8 files changed, 25 insertions(+), 43 deletions(-) > -- Siemens AG, Corporate Technology, CT RTC ITP SDP-DE Corporate Competence Center Embedded Linux