[Qemu-devel] [Bug 1791763] [NEW] broken signal handling in nios2 user-mode emulation

[Qemu-devel] [Bug 1791763] [NEW] broken signal handling in nios2 user-mode emulation

[Sandra Loosemore]

Public bug reported:

This bug is against the 3.0 release.

It appears that the signal handling parts of the nios2 user-mode
emulation have never really been completed or tested.  Some examples of
failing tests from the GCC testsuite are gcc.dg/pr78185.c and
gcc.dg/cleanup-10.c.

Some problems I've identified and tried to fix with the attached patch
are:

- Code copied from the Linux kernel wasn't adjusted to differentiate
between host and target data types and address spaces.

- The sigaltstack() system call returns EINVAL because fields are listed
in the wrong order in struct target_sigaltstack.

With this patch, the system calls to set up the signal handler are
returning successfully, but the handler isn't being invoked, so
something is still wrong.  I think I need another pair of eyes to look
over this code.

** Affects: qemu
     Importance: Undecided
         Status: New

** Patch added: "partial patch"
   https://bugs.launchpad.net/bugs/1791763/+attachment/5187164/+files/wip-qemu.patch

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1791763

Title:
  broken signal handling in nios2 user-mode emulation

Status in QEMU:
  New

Bug description:
  This bug is against the 3.0 release.

  It appears that the signal handling parts of the nios2 user-mode
  emulation have never really been completed or tested.  Some examples
  of failing tests from the GCC testsuite are gcc.dg/pr78185.c and
  gcc.dg/cleanup-10.c.

  Some problems I've identified and tried to fix with the attached patch
  are:

  - Code copied from the Linux kernel wasn't adjusted to differentiate
  between host and target data types and address spaces.

  - The sigaltstack() system call returns EINVAL because fields are
  listed in the wrong order in struct target_sigaltstack.

  With this patch, the system calls to set up the signal handler are
  returning successfully, but the handler isn't being invoked, so
  something is still wrong.  I think I need another pair of eyes to look
  over this code.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1791763/+subscriptions

[Bug 1791763] Re: broken signal handling in nios2 user-mode emulation

[Thomas Huth]

The QEMU project is currently considering to move its bug tracking to another system. For this we need to know which bugs are still valid and which could be closed already. Thus we are setting older bugs to "Incomplete" now.
If you still think this bug report here is valid, then please switch the state back to "New" within the next 60 days, otherwise this report will be marked as "Expired". Or mark it as "Fix Released" if the problem has been solved with a newer version of QEMU already. Thank you and sorry for the inconvenience.

** Changed in: qemu
       Status: New => Incomplete

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1791763

Title:
  broken signal handling in nios2 user-mode emulation

Status in QEMU:
  Incomplete

Bug description:
  This bug is against the 3.0 release.

  It appears that the signal handling parts of the nios2 user-mode
  emulation have never really been completed or tested.  Some examples
  of failing tests from the GCC testsuite are gcc.dg/pr78185.c and
  gcc.dg/cleanup-10.c.

  Some problems I've identified and tried to fix with the attached patch
  are:

  - Code copied from the Linux kernel wasn't adjusted to differentiate
  between host and target data types and address spaces.

  - The sigaltstack() system call returns EINVAL because fields are
  listed in the wrong order in struct target_sigaltstack.

  With this patch, the system calls to set up the signal handler are
  returning successfully, but the handler isn't being invoked, so
  something is still wrong.  I think I need another pair of eyes to look
  over this code.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1791763/+subscriptions

[Bug 1791763] Re: broken signal handling in nios2 user-mode emulation

[Peter Maydell]

A quick eyeball of the patch and the current QEMU tree indicates that at
least some of the bugs it's trying to fix still exist (notably a lot of
use of "long" in various target_* structures, which should not be using
types with a width dependent on the host system.)

** Changed in: qemu
       Status: Incomplete => Confirmed

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1791763

Title:
  broken signal handling in nios2 user-mode emulation

Status in QEMU:
  Confirmed

Bug description:
  This bug is against the 3.0 release.

  It appears that the signal handling parts of the nios2 user-mode
  emulation have never really been completed or tested.  Some examples
  of failing tests from the GCC testsuite are gcc.dg/pr78185.c and
  gcc.dg/cleanup-10.c.

  Some problems I've identified and tried to fix with the attached patch
  are:

  - Code copied from the Linux kernel wasn't adjusted to differentiate
  between host and target data types and address spaces.

  - The sigaltstack() system call returns EINVAL because fields are
  listed in the wrong order in struct target_sigaltstack.

  With this patch, the system calls to set up the signal handler are
  returning successfully, but the handler isn't being invoked, so
  something is still wrong.  I think I need another pair of eyes to look
  over this code.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1791763/+subscriptions

[Bug 1791763] Re: broken signal handling in nios2 user-mode emulation

[Thomas Huth]

This is an automated cleanup. This bug report has been moved to QEMU's
new bug tracker on gitlab.com and thus gets marked as 'expired' now.
Please continue with the discussion here:

 https://gitlab.com/qemu-project/qemu/-/issues/261


** Changed in: qemu
       Status: Confirmed => Expired

** Bug watch added: gitlab.com/qemu-project/qemu/-/issues #261
   https://gitlab.com/qemu-project/qemu/-/issues/261

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1791763

Title:
  broken signal handling in nios2 user-mode emulation

Status in QEMU:
  Expired

Bug description:
  This bug is against the 3.0 release.

  It appears that the signal handling parts of the nios2 user-mode
  emulation have never really been completed or tested.  Some examples
  of failing tests from the GCC testsuite are gcc.dg/pr78185.c and
  gcc.dg/cleanup-10.c.

  Some problems I've identified and tried to fix with the attached patch
  are:

  - Code copied from the Linux kernel wasn't adjusted to differentiate
  between host and target data types and address spaces.

  - The sigaltstack() system call returns EINVAL because fields are
  listed in the wrong order in struct target_sigaltstack.

  With this patch, the system calls to set up the signal handler are
  returning successfully, but the handler isn't being invoked, so
  something is still wrong.  I think I need another pair of eyes to look
  over this code.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1791763/+subscriptions