Bug#960634: nanopolish: FTBFS on ppc64el (as well as ppc64 and powerpc), enumerator conflict

[Barry Arndt]

Source: nanopolish
Version: 0.13.2-1
Severity: normal
Tags: ftbfs
Justification: fails to build from source (but built successfully in the 
past) on arch ppc64el
 
-- System Information:
Debian Release: bullseye/sid
  APT prefers unstable
  APT policy: (500, 'unstable')
Architecture: ppc64el (ppc64le)
Kernel: Linux 5.4.0-3-powerpc64le (SMP w/80 CPU cores)

nanopolish FTBFS on ppc64el:
https://buildd.debian.org/status/fetch.php?pkg=nanopolish=ppc64el=0.13.2-1=1588529870=0
 



nanopolish-0.13.2/src/nanopolish_squiggle_read.h contains the following 
enumeration:
 
enum PoreType
{
PT_UNKNOWN = 0,
PT_R7,
PT_R9,
};
 
Enumerators PT_R7 and PT_R9 are both already defined elsewhere, which 
conflicts with this package and causes FTBFS in ppc64el.  (I'm not sure 
where PT_R7 and PT_R9 are already defined.  A quick grep of kernel source 
shows several in /usr/src/linux-xx/arch as well as /usr/lib.)
 
I verified that changing those names (PT_R7 and PT_R9) in the enumeration 
and corresponding code fixes the problem in ppc64el and allows it to build 
successfully.  Based on the build log files for ppc64 and powerpc, this 
change should fix FTBFS for those as well. 
 
I did not include a patch.  Since this problem could be fixed in several 
different ways, including the possibility of changing enumerator names, 
the maintainer might want it to match already established conventions.
 
A minor nit, removing the comma after PT_R9 would make the code a little 
more standard and match the rest of the enums in that .h file.
 
Barry Arndt
 

Bug#866122: kernel resolution

[Barry Arndt]

For reference:
 
While debugging this very tricky problem, our kernel team found 2 separate 
but related bugs whose resolutions fix the problem outlined in this 
report.
 
The bug fixes both resulted in CVEs and have already been added to Linus' 
tree.  They are:
https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git/commit/?h=fixes=a8318c13e79badb92bc6640704a64cc022a6eb97
https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git/commit/?h=fixes=8205d5d98ef7f155de211f5e2eb6ca03d95a5a60
 
The CVEs are:
https://security-tracker.debian.org/tracker/CVE-2019-15030
https://security-tracker.debian.org/tracker/CVE-2019-15031
 
One of the problems was introduced in 4.12, and the other in 4.15.
 
Barry

Bug#866122: slapd-mtread crash on ppc64{,el} in stretch/sid

[Barry Arndt]

A few things:
 
I verified that this problem does not recreate on buster, as Ryan 
mentioned earlier, because of the disabling of lock elision in glibc.
 
I verified that this problem still exists on a Debian 9.8 base with 
upstream kernel 5.1.0-x, which contains up-to-date TM fixes.
 
Our kernel team is actively debugging this problem on upstream kernel 
5.2.0-rc5, and is in favor of disabling TM in Debian until this issue is 
resolved.
 
Thanks.
Barry

Bug#896843: FTBFS: incorrect CPPFLAG for powerpc*

[Barry Arndt]

Source: ctsim
Version: 6.0.2
Severity: serious
Tags: l10n patch
Justification: fails to build from source (but built successfully in the past)

Dear Maintainer,

FTBFS of ctsim-6.0.2 on ppc64el reported from unicamp.

The build log reported:
g++: error: unrecognized command line option '-faltivec'; did you mean 
'-maltivec'

The correct flag for ppc64el (as well as powerpc*) is -maltivec.

I changed -faltivec in configure.ac to -maltivec as follows:

--- ctsim-6.0.2.orig/configure.ac
+++ ctsim-6.0.2/configure.ac
@@ -137,7 +137,7 @@ case $target_cpu in
 CXXFLAGS="$CXXFLAGS $CPUEXT_FLAGS $SIMD_FLAGS"
 ;;
 powerpc*)
-ARCH_OPTION="-fno-common -faltivec";;
+ARCH_OPTION="-fno-common -maltivec";;
 armv1*|armv2*|armv3*|armv4*|armv5*|armv6*)
 ARCH_OPTION="-ffast-math";;
 armv7*|armv8*)

I then ran autoconf.  After that, the package build properly.

This fix should also fix the FTBFS of ctsim on 
powerpc, powerpcspe, and ppc64.



-- System Information:
Debian Release: buster/sid
  APT prefers unstable
  APT policy: (500, 'unstable')
Architecture: ppc64el (ppc64le)

Kernel: Linux 4.9.0-4-powerpc64le (SMP w/80 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), 
LANGUAGE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Bug#870909: RFS: cxlflash/4.3.2493-1 [ITP] -- IBM Data Engine for NoSQL Software Libraries

[Barry Arndt]

Mentors,

I am picking up work on this package.  I'm working through the issues 
already outlined to make sure they are resolved and all questions are 
answered.

Once that is done, I'll refresh the package if necessary, and inform of 
the updates and answered questions.

Thanks.

Barry Arndt
IBM Linux Technology Center - Linux on Power

Bug#866122: slapd-mtread crash on ppc64{,el} in stretch/sid

[Barry Arndt]

On Thu, 27 Jul 2017 10:20:12 -0300 Breno Leitao  
wrote:
> Although lack of recent updates, we are still working on this problem.
> 
> Barry (on CC) is allocated to work on this issue and should have updates 
soon.
> 
> 
> 

The offending line of code that Breno mentioned earlier was part of a 
larger patch.  Looking at that larger patch, I'm afraid that just removing 
the offending line of code would make that patch incomplete and could 
possibly leave other errors behind.

We have a few patches for other issues in this area of code.  I tested a 
couple of them, but they did not resolve this issue.  I'm working with 
another developer (who is currently working on other items related to this 
one) to gather more debug information to narrow down to the correct fix.