From: Ulrich Weigand <ulrich.weigand(a)linaro.org>
GDB's interrupt.exp test cases currenly fail on ARM. The problem is how do_signal
handled restarting interrupted system calls:
The entry.S assembler code determines that we come from a system call; and that
information is passed as "syscall" parameter to do_signal. That routine then
calls get_signal_to_deliver [*] and if a signal is to be delivered, calls into
handle_signal. If a system call is to be restarted either after the signal
handler returns, or if no handler is to be called in the first place, the PC
is updated after the get_signal_to_deliver call, either in handle_signal (if
we have a handler) or at the end of do_signal (otherwise).
Now the problem is that during [*], the call to get_signal_to_deliver, a ptrace
intercept may happen. During this intercept, the debugger may change registers,
including the PC. This is done by GDB if it wants to execute an "inferior call",
i.e. the execution of some code in the debugged program triggered by GDB.
To this purpose, GDB will save all registers, allocate a stack frame, set up
PC and arguments as appropriate for the call, and point the link register to
a dummy breakpoint instruction. Once the process is restarted, it will execute
the call and then trap back to the debugger, at which point GDB will restore
all registers and continue original execution.
This generally works fine. However, now consider what happens when GDB attempts
to do exactly that while the process was interrupted during execution of a to-be-
restarted system call: do_signal is called with the syscall flag set; it calls
get_signal_to_deliver, at which point the debugger takes over and changes the PC
to point to a completely different place. Now get_signal_to_deliver returns
without a signal to deliver; but now do_signal decides it should be restarting
a system call, and decrements the PC by 2 or 4 -- so it now points to 2 or 4
bytes before the function GDB wants to call -- which leads to a subsequent crash.
To fix this problem, two things need to be supported:
- do_signal must be able to recognize that get_signal_to_deliver changed the PC
to a different location, and skip the restart-syscall sequence
- once the debugger has restored all registers at the end of the inferior call
sequence, do_signal must recognize that *now* it needs to restart the pending
system call, even though it was now entered from a breakpoint instead of an
actual svc instruction
This set of issues is solved on other platforms, usually by one of two
mechanisms:
- The status information "do_signal is handling a system call that may need
restarting" is itself carried in some register that can be accessed via
ptrace. This is e.g. on Intel the "orig_eax" register; on Sparc the kernel
defines a magic extra bit in the flags register for this purpose.
This allows GDB to manage that state: reset it when doing an inferior call,
and restore it after the call is finished.
- On s390, do_signal transparently handles this problem without requiring
GDB interaction, by performing system call restarting in the following
way: first, adjust the PC as necessary for restarting the call. Then,
call get_signal_to_deliver; and finally just continue execution at the
PC. This way, if GDB does not change the PC, everything is as before.
If GDB *does* change the PC, execution will simply continue there --
and once GDB restores the PC it saved at that point, it will automatically
point to the *restarted* system call. (There is the minor twist how to
handle system calls that do *not* need restarting -- do_signal will undo
the PC change in this case, after get_signal_to_deliver has returned, and
only if ptrace did not change the PC during that call.)
Because there does not appear to be any obvious register to carry the
syscall-restart information on ARM, we'd either have to introduce a new
artificial ptrace register just for that purpose, or else handle the issue
transparently like on s390. The patch below implements the second option;
using this patch makes the interrupt.exp test cases pass on ARM, with no
regression in the GDB test suite otherwise.
Signed-off-by: Ulrich Weigand <ulrich.weigand(a)linaro.org>
Signed-off-by: Arnd Bergmann <arnd.bergmann(a)linaro.org>
Cc: patches(a)linaro.org
---
arch/arm/kernel/signal.c | 90 +++++++++++++++++++++++++++++++++++++--------------------------
1 file changed, 53 insertions(+), 37 deletions(-)
diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c
index cb83983..8242197 100644
--- a/arch/arm/kernel/signal.c
+++ b/arch/arm/kernel/signal.c
@@ -597,19 +597,13 @@ setup_rt_frame(int usig, struct k_sigaction *ka, siginfo_t *info,
return err;
}
-static inline void setup_syscall_restart(struct pt_regs *regs)
-{
- regs->ARM_r0 = regs->ARM_ORIG_r0;
- regs->ARM_pc -= thumb_mode(regs) ? 2 : 4;
-}
-
/*
* OK, we're invoking a handler
*/
static int
handle_signal(unsigned long sig, struct k_sigaction *ka,
siginfo_t *info, sigset_t *oldset,
- struct pt_regs * regs, int syscall)
+ struct pt_regs * regs)
{
struct thread_info *thread = current_thread_info();
struct task_struct *tsk = current;
@@ -617,26 +611,6 @@ handle_signal(unsigned long sig, struct k_sigaction *ka,
int ret;
/*
- * If we were from a system call, check for system call restarting...
- */
- if (syscall) {
- switch (regs->ARM_r0) {
- case -ERESTART_RESTARTBLOCK:
- case -ERESTARTNOHAND:
- regs->ARM_r0 = -EINTR;
- break;
- case -ERESTARTSYS:
- if (!(ka->sa.sa_flags & SA_RESTART)) {
- regs->ARM_r0 = -EINTR;
- break;
- }
- /* fallthrough */
- case -ERESTARTNOINTR:
- setup_syscall_restart(regs);
- }
- }
-
- /*
* translate the signal
*/
if (usig < 32 && thread->exec_domain && thread->exec_domain->signal_invmap)
@@ -685,6 +659,7 @@ handle_signal(unsigned long sig, struct k_sigaction *ka,
*/
static void do_signal(struct pt_regs *regs, int syscall)
{
+ unsigned int retval = 0, continue_addr = 0, restart_addr = 0;
struct k_sigaction ka;
siginfo_t info;
int signr;
@@ -698,18 +673,61 @@ static void do_signal(struct pt_regs *regs, int syscall)
if (!user_mode(regs))
return;
+ /*
+ * If we were from a system call, check for system call restarting...
+ */
+ if (syscall) {
+ continue_addr = regs->ARM_pc;
+ restart_addr = continue_addr - (thumb_mode(regs) ? 2 : 4);
+ retval = regs->ARM_r0;
+
+ /*
+ * Prepare for system call restart. We do this here so that a
+ * debugger will see the already changed PSW.
+ */
+ switch (retval) {
+ case -ERESTARTNOHAND:
+ case -ERESTARTSYS:
+ case -ERESTARTNOINTR:
+ regs->ARM_r0 = regs->ARM_ORIG_r0;
+ regs->ARM_pc = restart_addr;
+ break;
+ case -ERESTART_RESTARTBLOCK:
+ regs->ARM_r0 = -EINTR;
+ break;
+ }
+ }
+
if (try_to_freeze())
goto no_signal;
+ /*
+ * Get the signal to deliver. When running under ptrace, at this
+ * point the debugger may change all our registers ...
+ */
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
sigset_t *oldset;
+ /*
+ * Depending on the signal settings we may need to revert the
+ * decision to restart the system call. But skip this if a
+ * debugger has chosen to restart at a different PC.
+ */
+ if (regs->ARM_pc == restart_addr) {
+ if (retval == -ERESTARTNOHAND
+ || (retval == -ERESTARTSYS
+ && !(ka.sa.sa_flags & SA_RESTART))) {
+ regs->ARM_r0 = -EINTR;
+ regs->ARM_pc = continue_addr;
+ }
+ }
+
if (test_thread_flag(TIF_RESTORE_SIGMASK))
oldset = ¤t->saved_sigmask;
else
oldset = ¤t->blocked;
- if (handle_signal(signr, &ka, &info, oldset, regs, syscall) == 0) {
+ if (handle_signal(signr, &ka, &info, oldset, regs) == 0) {
/*
* A signal was successfully delivered; the saved
* sigmask will have been stored in the signal frame,
@@ -723,11 +741,14 @@ static void do_signal(struct pt_regs *regs, int syscall)
}
no_signal:
- /*
- * No signal to deliver to the process - restart the syscall.
- */
if (syscall) {
- if (regs->ARM_r0 == -ERESTART_RESTARTBLOCK) {
+ /*
+ * Handle restarting a different system call. As above,
+ * if a debugger has chosen to restart at a different PC,
+ * ignore the restart.
+ */
+ if (retval == -ERESTART_RESTARTBLOCK
+ && regs->ARM_pc == continue_addr) {
if (thumb_mode(regs)) {
regs->ARM_r7 = __NR_restart_syscall - __NR_SYSCALL_BASE;
regs->ARM_pc -= 2;
@@ -750,11 +771,6 @@ static void do_signal(struct pt_regs *regs, int syscall)
#endif
}
}
- if (regs->ARM_r0 == -ERESTARTNOHAND ||
- regs->ARM_r0 == -ERESTARTSYS ||
- regs->ARM_r0 == -ERESTARTNOINTR) {
- setup_syscall_restart(regs);
- }
/* If there's no signal to deliver, we just put the saved sigmask
* back.
Last week a new tab was added to the banner of wiki.linaro.org
called "HowTo". This links to a page we have been working on
that is intended to point people to some of our most common
how-to scenarios. In addition it includes a pointer to pages
that are tagged with "CategoryHowTo", just click on the "list
of all how-to's" link. We hope that promoting the visibility
of this page will make it easier for others to find the Linaro
information they are looking for.
Please take a look at the page when you get some time:
https://wiki.linaro.org/Resources/HowTo
There might be a link you'll find helpful.
Also, please let us know if you think there is an existing page
in the wiki that we should add to this page.
Hello Everybody,
The Engineering Resources team weekly report for 2011-3-17 can be
found at:
https://wiki.linaro.org/Resources/Status/WeeklyMeeting/2011-03-17
Highlights below.
Best regards,
Matt
=== Action Items from This Meeting ===
* Improve the Linaro planning structure guides. ie:
* engineering BP versus TR BP, when to write a spec
* what's a good quality BP
* Linaro at UDS introduction for new employees
* what are sessions supposed to be about? you need to know what
you are going to do *before* you show up. but not too far.
* what to expect at the summit
* Firewall issues. Use VM's to hack around some of the issues
* IRC setup help
=== Action Items from Last Meeting ===
* Action items from last meeting with tech leads
* AddNewHardware and HardwarePack pages need to explain kernel
packages as Steve noted
* HardwarePacks page refers to Maverick in stead of Natty - Done
* Mentoring section needs to be renamed and back links should be
fixed - Done
* A better header for wiki that includes things like a "Resources"
link in banner/header - Done
* Toolchain documentation improvements
I have this Integrator/AP development board here in my office:
Initialising Boot Monitor System Information Block
boot Monitor > i
ARM bootPROM [Version 1.3] Rebuilt on Jun 26 2001 at 22:04:10
Running on a Integrator Evaluation Board
Board Revision V1.0, ARM920T Processor
Memory Size is 128MBytes, Flash Size is 32MBytes
Copyright (c) ARM Limited 1999 - 2001. All rights reserved.
Board designed by ARM Limited
Hardware support provided at http://www.arm.com/
For help on the available commands type ? or h
boot Monitor >
Does anyone beside me *ever* try to actually boot a recent kernel
on this thing?
How do you do it? Do I need patched U-boots etc, or can I just load
a Z-record image of vmlinux into RAM over the serial console?
Needless to say, ARM Ltd. has deleted the web pages for official
support of this thing as far as I can see. I was going to try out a
few patches affecting mach-integrator...
Yours,
Linus Walleij
>
> Hi All,
>
> Please find enclosed link to minutes and actions for multimedia wg
> meeting on 15st March 2011.
>
>
> https://wiki.linaro.org/WorkingGroups/Middleware/Multimedia/Notes/2011-03-15
>
> Summary
> - Kan created seperate AAC encoder lib from Android aac encoder.Talking to
> sourceforge opencore-amr maintainers to see if they are willing to maintain
> opencore-aac encoder as well
>
- Bufferpool implementation in xvimagesink patch sent for review to Wim.No
> feedback yet
> - System metrics app and integrated instrumented player packaging on going
>
>
> Thanks
> Sachin
>
This patch set is to take sdhci device driver specific things out
from sdhci-pltfm.c and make them self registered. Here are the
difference it makes.
* Get the sdhci device driver follow the Linux trend that driver
take the registration by its own
* sdhci-pltfm.c becomes significantly simpler and only has common
support functions there
* All sdhci device specific stuff are going back its own driver
* The dt and non-dt support share the same pair of .probe and
.remove hooks.
The first one patch adds common support function into sdhci-pltfm.c
when changing sdhci-esdhc-imx driver, while the last one clean up
sdhci-pltfm.c when changing sdhci-tegra driver.
Only the patch of sdhci-esdhc-imx are tested on hardware, and others
are just build tested.
Regards,
Shawn
Shawn Guo (5):
mmc: sdhci: make sdhci-esdhc-imx driver self registered
mmc: sdhci: make sdhci-cns3xxx driver self registered
mmc: sdhci: make sdhci-dove driver self registered
mmc: sdhci: make sdhci-tegra driver self registered
mmc: sdhci: update Makefile/Kconfig for sdhci_pltfm change
drivers/mmc/host/Kconfig | 24 +++--
drivers/mmc/host/Makefile | 11 +-
drivers/mmc/host/sdhci-cns3xxx.c | 68 +++++++++++++-
drivers/mmc/host/sdhci-dove.c | 70 +++++++++++++-
drivers/mmc/host/sdhci-esdhc-imx.c | 98 +++++++++++++++----
drivers/mmc/host/sdhci-pltfm.c | 165 ++-----------------------------
drivers/mmc/host/sdhci-pltfm.h | 14 ++-
drivers/mmc/host/sdhci-tegra.c | 189 +++++++++++++++++++++---------------
Hello all,
I'm looking at function of_device_alloc in drivers/of/platform.c, and
surprisingly found that the platform device instance id is being
hard-coded as -1 when calling platform_device_alloc. Does this mean
that there is no support of pdev id in DT? If yes, can anyone please
help me understand why we do not have something in DT node to reflect
this id and get it supported? Or this is something on TODO list?
I'm asking this because I feel we still need pdev id support in DT when
reviewing the patch set of 'mx51 basic DT support'.
--
Regards,
Shawn
