Bug#1055750: gfortran: [armhf] Yield SIGBUS when compiling with -fstack-clash-protection
* Emanuele Rocca [2023-11-14 12:01]: On 2023-11-13 05:13, Rafael Laboissière wrote: The attached file bug-1055750.tgz contains a minimal code that triggers the bug on an armhf system Thanks! For the record I can reproduce the issue in a armhf chroot, but *not* on armel and arm64. The only thing to change in the reproducer is the -I argument to gfortran: armhf: /usr/lib/arm-linux-gnueabihf armel: /usr/lib/arm-linux-gnueabi arm64: /usr/lib/aarch64-linux-gnu I'll see if I can find out more. Thanks for the followup, Emanuele. I have investigated the issue a little bit deeper. Please, find here attached a new tarball with a simplified version of the x09f.f90 file, that still triggers the bug. Running the resulting executable with gbd, yields the following: (gdb) run -dev ps -o /dev/null Starting program: /home/rafael/bug-1055750/x09f -dev ps -o /dev/null [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1". Program received signal SIGSEGV, Segmentation fault. x09f::mypltr2 (x=0, y=1, xt=0, yt=7.300765e-43) at x09f.f90:48 48 yt = tr(1) In this new code, there are two private functions mypltr1 and mypltr2. The bug only happens when the latter is invoked, in which a value coming from the global variable tr is assigned to yt. In mypltr1, the assignment is done to xt, instead of yt, and no segfault is triggered. Best, Rafael Laboissière bug-1055750.tgz Description: application/gtar-compressed
Bug#1055750: gfortran: [armhf] Yield SIGBUS when compiling with -fstack-clash-protection
* Rafael Laboissière [2023-11-15 18:06]: I have investigated the issue a little bit deeper. Please, find here attached a new tarball with a simplified version of the x09f.f90 file, that still triggers the bug. Running the resulting executable with gbd, yields the following: Oops, with the correct tarball this time. Best, Rafael Laboissière bug-1055750.tgz Description: application/gtar-compressed
Bug#1055750: gfortran: [armhf] Yield SIGBUS when compiling with -fstack-clash-protection
* Rafael Laboissière [2023-11-15 18:37]: * Rafael Laboissière [2023-11-15 18:06]: I have investigated the issue a little bit deeper. Please, find here attached a new tarball with a simplified version of the x09f.f90 file, that still triggers the bug. Running the resulting executable with gbd, yields the following: Oops, with the correct tarball this time. Grrr, with the correct one now, hopefully! R. bug-1055750.tgz Description: application/gtar-compressed
Bug#1055750: Bug#1055228: plplot: FTBFS on armhf (test segfault)
Hi Emanuele, Our messages crossed. * Emanuele Rocca [2023-11-15 18:20]: On 2023-11-09 05:11, Rafael Laboissière wrote: The Fortran example x09f.f90, which is exercised during the building of plplot, now fails on armhf, due to the use of the compiler option -fstack-clash-protection. The problem seems unrelated to stach-clash-protection I think, enabling the feature on armhf just made it evident. Building the program on a x86 system without -fstack-clash-protection but with -fsanitize=address, it segfaults: /usr/bin/gfortran -g -O2 x09f.f90 -o x09f -I/usr/lib/x86_64-linux-gnu/fortran/modules/plplot -lplplotfortran -fsanitize=address ./x09f -dev ps -o /dev/null Program received signal SIGSEGV: Segmentation fault - invalid memory reference. The SIGSEGV happens in plplot_single::pltransformf2c, which is about where the SIGBUS happens on armhf with stack-clash-protector enabled: Program received signal SIGSEGV, Segmentation fault. 0x75a000a8 in ?? () (gdb) bt #0 0x75a000a8 in ?? () #1 0x77f05683 in plplot_single::pltransformf2c (x=x@entry=0, y=y@entry=1, tx=4.9406564584124654e-324, ty=6.9533558054215925e-310, data=) at ./bindings/fortran/plplot_single.f90:114 Here's the SIGBUS on armhf with stack-clash-protection: Program received signal SIGBUS, Bus error. 0x00400822 in x09f::mypltr (x=0, y=1, xt=0, yt=0) at x09f.f90:38 38 xt = tr(1) * x + tr(2) * y + tr(3) (gdb) bt #0 0x00400822 in x09f::mypltr (x=0, y=1, xt=0, yt=0) at x09f.f90:38 #1 0xf7f58a06 in plplot_single::pltransformf2c (x=, y=, tx=-9.8841854819221187e+269, ty=-nan(0xe5db40001), data=) at ./bindings/fortran/plplot_single.f90:114 Does this mean that the origin of the bug is upstream or that it still may be a bug in gfortran? Best, Rafael Laboissière
Bug#1055750: Bug#1055228: plplot: FTBFS on armhf (test segfault)
Hello Rafael! On 2023-11-15 06:47, Rafael Laboissière wrote: > Does this mean that the origin of the bug is upstream or that it still may > be a bug in gfortran? At this point we know for sure that the issue is not armhf-specific, and also that it is not caused by stack-clash-protection. On the contrary, enabling stack-clash-protection on armhf allowed us to discover a problem that would have otherwise gone unnoticed. Whether the bug is in plplot or gfortran I really have no idea. I would tend not to think of a compiler issue unless we have some evidence in that direction though, and suggest raising the issue with plplot upstream. Showing them the x86 reproducer with -fsanitize=address should be a good starting point. Thanks, Emanuele
Bug#1055750: Bug#1055228: plplot: FTBFS on armhf (test segfault)
* Emanuele Rocca [2023-11-15 20:11]: On 2023-11-15 06:47, Rafael Laboissière wrote: Does this mean that the origin of the bug is upstream or that it still may be a bug in gfortran? At this point we know for sure that the issue is not armhf-specific, and also that it is not caused by stack-clash-protection. On the contrary, enabling stack-clash-protection on armhf allowed us to discover a problem that would have otherwise gone unnoticed. Whether the bug is in plplot or gfortran I really have no idea. I would tend not to think of a compiler issue unless we have some evidence in that direction though, and suggest raising the issue with plplot upstream. Showing them the x86 reproducer with -fsanitize=address should be a good starting point. Ok, thanks. FYI, I can reproduce the segfault on armhf and amd64 with -fsanitize=address. My guess is that the bug is in PLplot and not in gfortran, but this is jsut a guess. I will eventually inform the PLplot upstream authors about the issue. Best, Rafael Laboissière
Re: gcc-multilib & g++-multilib issues
Please stop filing duplicating bogus reports. You make unproven claims without giving any rationale. On 14.11.23 00:04, jahway-bug-reports wrote: Hi Matthias, There is software which will not build natively on Debian 10/11/12 without the gcc-multilib & g++-multilib packages https://packages.debian.org/bookworm/gcc-multilib https://packages.debian.org/bookworm/g++-multilib I can build it perfectly fine on amd64, but these packages are missing for me to build natively on a ARM 64 platform (such as a Raspberry Pi). I have already filed the following bugs [1][2][3][4] to include these packages, but you give no explanation other than "arm64 doesn't have multilibs" but it would if these bugs were explored and resolved so that these packages would be available on the arm64 platform. Can you please tell me why the Debian project is not able to consider adding these packages to ARM64 so that software can be natively compiled on this platform running Debian? Thanks [1] March 17, 2023 - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1033145 [2] March 17, 2023 - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1033146 [3] Nov 1, 2023 - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1055199 [4] Nov 1, 2023 - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1055200
Processed: Re: Bug#1055228: plplot: FTBFS on armhf (test segfault)
Processing control commands: > forwarded -1 https://sourceforge.net/p/plplot/bugs/206/ Bug #1055750 [gfortran] gfortran: [armhf] Yield SIGBUS when compiling with -fstack-clash-protection Set Bug forwarded-to-address to 'https://sourceforge.net/p/plplot/bugs/206/'. -- 1055750: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1055750 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Bug#1055750: Bug#1055228: plplot: FTBFS on armhf (test segfault)
Control: forwarded -1 https://sourceforge.net/p/plplot/bugs/206/ * Rafael Laboissière [2023-11-16 07:51]: My guess is that the bug is in PLplot and not in gfortran, but this is just a guess. I will eventually inform the PLplot upstream authors about the issue. Done ! R.
