all messages for Guix-related lists mirrored at yhetil.org
 help / color / mirror / code / Atom feed
* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
@ 2022-01-16  6:12 elaexuotee--- via Bug reports for GNU Guix
  2022-01-17 15:42 ` Ludovic Courtès
  0 siblings, 1 reply; 7+ messages in thread
From: elaexuotee--- via Bug reports for GNU Guix @ 2022-01-16  6:12 UTC (permalink / raw)
  To: 53296

[-- Attachment #1: Type: text/plain, Size: 9110 bytes --]

Hey Guix,

Recently working on a package, I have been encounting intermittent segfault
during a build.

The segfault seems to only occur when I have some error in my code that causes
a crash, and the segfaults tend to cluster, appearing unexpectedly for a few
build attempts, and then disappearing right as I think they are reproducible
and try to grab an strace or something. Unfortunately, that's about the extent
of information I have been able to gleen.

The latest segfault happened with the attached package definition. Note, the
offending code is at line 77, where I forgot to remove a docstring from a
variable that used to be a procedure.

Given that non-segfault runs seem to error out so early, is this better thought
to be an issue with Guile? FWIW, I see nothing interesting under
/var/log/guix-daemon.log. However, lines like following show up in the kernel
messages ring:

[318026.268095] guix[7419]: segfault at 18 ip 00007f56ef6a01a3 sp 00007fff15588980 error 4 in libgc.so.1.4.3[7f56ef693000+1b000]
[318026.268116] Code: 8d 2d 71 93 01 00 90 4a 8d 04 e5 00 00 00 00 48 89 04 24 49 8b 45 00 4e 8b 3c e0 4d 85 ff 74 2a 31 ed 0f 1f 44 00 00 4d 89 fe <4d> 8b 7f 08 49 8b 7e 10 48 f7 d7 e8 6d 35 ff ff 85 c0 0f 84 3d 01
[318029.715621] guix[7761]: segfault at 10 ip 00007f9e80b919b9 sp 00007fffd1b2ad20 error 4 in libgc.so.1.4.3[7f9e80b7b000+1b000]
[318029.715638] Code: f7 d2 48 21 d0 48 8b 13 4c 8d 3c c5 00 00 00 00 48 8b 04 c2 48 85 c0 74 78 48 89 ea 48 f7 d2 eb 09 48 8b 40 08 48 85 c0 74 67 <48> 39 10 75 f2 44 8b 05 03 1c 04 00 49 f7 d4 4c 89 60 10 41 bc 01
[318041.537171] guix[8660]: segfault at 10 ip 00007f0d2603c9b9 sp 00007ffc72e998d0 error 4 in libgc.so.1.4.3[7f0d26026000+1b000]
[318041.537185] Code: f7 d2 48 21 d0 48 8b 13 4c 8d 3c c5 00 00 00 00 48 8b 04 c2 48 85 c0 74 78 48 89 ea 48 f7 d2 eb 09 48 8b 40 08 48 85 c0 74 67 <48> 39 10 75 f2 44 8b 05 03 1c 04 00 49 f7 d4 4c 89 60 10 41 bc 01


In the off chance it's helpful, below are some random machine details. Please
let me know if there is anything more pointed or specific I provide.

$ guix system describe

Generation 4	Jan 12 2022 18:59:48	(current)
  file name: /var/guix/profiles/system-4-link
  canonical file name: /gnu/store/sb01mnd31a9x2a0bznzlb2lsy91qwgk6-system
  label: GNU with Linux 5.15.13
  bootloader: grub-efi
  root device: label: "root"
  kernel: /gnu/store/bdf2yw10jr02mhyiwm05yp2qibywqz47-linux-5.15.13/bzImage
  channels:
    guix-bmw:
      repository URL: git://git@git.wilsonb.com/guix-bmw.git
      branch: master
      commit: 9fb59483371bb5d59fbd27e47baac88263410ac5
    nonguix:
      repository URL: https://gitlab.com/nonguix/nonguix
      branch: master
      commit: 023508df4804dbd9f39cb197525f166bc259f995
    guix:
      repository URL: https://git.savannah.gnu.org/git/guix.git
      branch: master
      commit: 9a2cf2c9232e229f7bb1ab065df2cf0740f65996
  configuration file: /gnu/store/pkf4vzlck0g32hkyvijmlcnp15vh8njv-configuration.scm

$ lscpu
Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Address sizes:                   43 bits physical, 48 bits virtual
Byte Order:                      Little Endian
CPU(s):                          8
On-line CPU(s) list:             0-7
Vendor ID:                       AuthenticAMD
Model name:                      AMD Ryzen 5 3500U with Radeon Vega Mobile Gfx
CPU family:                      23
Model:                           24
Thread(s) per core:              2
Core(s) per socket:              4
Socket(s):                       1
Stepping:                        1
Frequency boost:                 enabled
CPU max MHz:                     2100.0000
CPU min MHz:                     1400.0000
BogoMIPS:                        4191.75
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb hw_pstate ssbd ibpb vmmcall fsgsbase bmi1 avx2 smep bmi2 rdseed adx smap clflushopt sha_ni xsaveopt xsavec xgetbv1 xsaves clzero irperf xsaveerptr arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif overflow_recov succor smca sme sev sev_es
Virtualization:                  AMD-V
L1d cache:                       128 KiB (4 instances)
L1i cache:                       256 KiB (4 instances)
L2 cache:                        2 MiB (4 instances)
L3 cache:                        4 MiB (1 instance)
NUMA node(s):                    1
NUMA node0 CPU(s):               0-7
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Not affected
Vulnerability Mds:               Not affected
Vulnerability Meltdown:          Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Full AMD retpoline, IBPB conditional, STIBP disabled, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Not affected

$ lsmem
RANGE                                 SIZE  STATE REMOVABLE  BLOCK
0x0000000000000000-0x00000000bfffffff   3G online       yes   0-23
0x0000000100000000-0x00000005bfffffff  19G online       yes 32-183

Memory block size:       128M
Total online memory:      22G
Total offline memory:      0B

$ lsirq
IRQ     TOTAL NAME
RES 535178549 Rescheduling interrupts
LOC 262669447 Local timer interrupts
 82 125390312 PCI-MSI 2621440-edge amdgpu
CAL  66644945 Function call interrupts
TLB  28850164 TLB shootdowns
 12   7316777 IO-APIC 12-edge i8042
 77   2140601 PCI-MSI 2097156-edge iwlwifi:queue_4
  9   1653508 IO-APIC 9-fasteoi acpi
 73   1583984 PCI-MSI 2097152-edge iwlwifi:default_queue
IWI   1102407 IRQ work interrupts
 65    759824 PCI-MSI 3145728-edge ahci[0000:06:00.0]
  1    532374 IO-APIC 1-edge i8042
 76    253951 PCI-MSI 2097155-edge iwlwifi:queue_3
 75    186427 PCI-MSI 2097154-edge iwlwifi:queue_2
 74    168179 PCI-MSI 2097153-edge iwlwifi:queue_1
  7    100000 IO-APIC 7-fasteoi pinctrl_amd
 52     41159 PCI-MSI 524289-edge nvme0q1
 58     38523 PCI-MSI 524295-edge nvme0q7
 56     38206 PCI-MSI 524293-edge nvme0q5
 53     36960 PCI-MSI 524290-edge nvme0q2
 55     30286 PCI-MSI 524292-edge nvme0q4
 54     28877 PCI-MSI 524291-edge nvme0q3
 57     27421 PCI-MSI 524294-edge nvme0q6
 59     27142 PCI-MSI 524296-edge nvme0q8
MCP      8512 Machine check polls
 35      2317 PCI-MSI 2627584-edge xhci_hcd
 81      1290 PCI-MSI 2633728-edge snd_hda_intel:card1
 44      1215 PCI-MSI 2629632-edge xhci_hcd
 67       338 PCI-MSI 1572864-edge rtsx_pci
 80       316 PCI-MSI 2623488-edge snd_hda_intel:card0
  0        34 IO-APIC 2-edge timer
 78        33 PCI-MSI 2097157-edge iwlwifi:exception
 34        28 PCI-MSI 524288-edge nvme0q0
  8         1 IO-APIC 8-edge rtc0
 25         0 PCI-MSI 18432-edge PCIe PME, aerdrv
 26         0 PCI-MSI 20480-edge PCIe PME, aerdrv
 27         0 PCI-MSI 22528-edge PCIe PME, aerdrv
 28         0 PCI-MSI 28672-edge PCIe PME, aerdrv, pciehp
 29         0 PCI-MSI 133120-edge PCIe PME
 30         0 PCI-MSI 135168-edge PCIe PME
 36         0 PCI-MSI 2627585-edge xhci_hcd
 37         0 PCI-MSI 2627586-edge xhci_hcd
 38         0 PCI-MSI 2627587-edge xhci_hcd
 39         0 PCI-MSI 2627588-edge xhci_hcd
 40         0 PCI-MSI 2627589-edge xhci_hcd
 41         0 PCI-MSI 2627590-edge xhci_hcd
 42         0 PCI-MSI 2627591-edge xhci_hcd
 45         0 PCI-MSI 2629633-edge xhci_hcd
 46         0 PCI-MSI 2629634-edge xhci_hcd
 47         0 PCI-MSI 2629635-edge xhci_hcd
 48         0 PCI-MSI 2629636-edge xhci_hcd
 49         0 PCI-MSI 2629637-edge xhci_hcd
 50         0 PCI-MSI 2629638-edge xhci_hcd
 51         0 PCI-MSI 2629639-edge xhci_hcd
 60         0 PCI-MSI 524297-edge nvme0q9
 61         0 PCI-MSI 524298-edge nvme0q10
 62         0 PCI-MSI 524299-edge nvme0q11
 63         0 PCI-MSI 524300-edge nvme0q12
 71         0 PCI-MSI 2625537-edge ccp-1
 79         0 PCI-MSI 1048576-edge enp2s0
NMI         0 Non-maskable interrupts
SPU         0 Spurious interrupts
PMI         0 Performance monitoring interrupts
RTR         0 APIC ICR read retries
TRM         0 Thermal event interrupts
THR         0 Threshold APIC interrupts
DFR         0 Deferred Error APIC interrupts
MCE         0 Machine check exceptions
ERR         0 
MIS         0 
PIN         0 Posted-interrupt notification event
NPI         0 Nested posted-interrupt event
PIW         0 Posted-interrupt wakeup event



[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: jsoftware.scm --]
[-- Type: text/x-patch, Size: 20883 bytes --]

;;; GNU Guix --- Functional package management for GNU
;;; Copyright © 2022 B. Wilson <elaexuotee@wilsonb.com>
;;;
;;; This file is part of GNU Guix.
;;;
;;; GNU Guix is free software; you can redistribute it and/or modify it
;;; under the terms of the GNU General Public License as published by
;;; the Free Software Foundation; either version 3 of the License, or (at
;;; your option) any later version.
;;;
;;; GNU Guix is distributed in the hope that it will be useful, but
;;; WITHOUT ANY WARRANTY; without even the implied warranty of
;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
;;; GNU General Public License for more details.
;;;
;;; You should have received a copy of the GNU General Public License
;;; along with GNU Guix.  If not, see <http://www.gnu.org/licenses/>.

(define-module (gnu packages jsoftware)
  #:use-module (guix build utils)
  #:use-module (guix build-system gnu)
  #:use-module (guix build-system trivial)
  #:use-module (guix git-download)
  #:use-module ((guix licenses) #:prefix license:)
  #:use-module (guix packages)
  #:use-module (guix utils)
  #:use-module (gnu packages)
  #:use-module (gnu packages libedit)
  #:use-module (gnu packages llvm)
  #:use-module (gnu packages maths)
  #:use-module (guix gexp)
  #:use-module (ice-9 ftw)
  #:use-module (ice-9 match)
  #:use-module (ice-9 regex)
  #:use-module (ice-9 rdelim)
  #:use-module (srfi srfi-1)
  #:use-module (srfi srfi-26)
  #:use-module (srfi srfi-71))


;;; TODO: Make importer and packages for J addons:
;;; http://www.jsoftware.com/jal/

;;; TODO: Package up j80x series


(define (jname prefix release-type)
  "Return a package name for J, including RELEASE-TYPE only if not 'release."
  (match release-type
    ('release prefix)
    (_        (string-append prefix "-" (symbol->string release-type)))))

;; We want a version string where packages specifications like pkg@MAJOR work.
;; This requires that the first version part separator be dot.  Subsequent
;; separators are hyphen, mirror `git-version' etc.
(define* (jversion->string major #:optional minor revision commit)
  "Return a version string formatted like MAJOR.MINOR-REVISION-COMMIT.  Only
  MAJOR is required, and MINOR defaults to ``0'' if not supplied."
  (let* ((commit (and commit (string-take commit 7)))
         (minor (or minor "0"))
         (sub-parts (filter (cut (compose not eq?) #f <>)
                            (list minor revision commit))))
    (string-append major "." (string-join sub-parts "-"))))

(define* (jrelease-string release-type #:optional version-minor)
  "Construct J release identifier string."
  (let ((release-type (symbol->string release-type)))
    (if version-minor
     (string-append release-type "-" version-minor)
     release-type)))

(define* (jinfo->git-tag version-major release-type #:optional version-minor)
  "Given version parameters, construct a git tag for upstream releases."
  (string-append "j" version-major (jrelease-string release-type version-minor)))

(define ijconsole
  "G-exp script that detects AVX/AVX2 support at runtime and executes jconsole
  with the appropriate libj.so and profile.ijs."
  (with-imported-modules '((guix cpu)
                           (guix memoization)
                           (guix profiling)
                           (guix sets)
                           (srfi srfi-26))
    (program-file "ijconsole"
      #~(begin
          (use-modules ((guix cpu)     #:select (cpu-flags current-cpu))
                       ((guix sets)    #:select (set-contains?))
                       ((srfi srfi-26) #:select (cute)))

          ;; Assume that this script will be installed under bin/.
          (define %basedir (dirname (dirname (current-filename))))

          (let* ((jconsole (string-append %basedir "/libexec/j/jconsole"))
                 (cpu-has-flag?
                   (cute set-contains? (cpu-flags (current-cpu)) <>))
                 (libj (format #f "~a/lib/j/libj~a.so" %basedir
                               (cond ((cpu-has-flag? "avx2") "-avx2")
                                     ((cpu-has-flag? "avx") "-avx")
                                     (else ""))))
                 (jprofile (string-append %basedir "/etc/j/profile.ijs")))
            (apply execl jconsole "ijconsole" "-lib" libj "-jprofile" jprofile
                   (cdr (command-line))))))))

(define* (make-j #:key
                 version
                 revision
                 hash
                 tag
                 commit
                 (release-type 'release)
                 (patches '())
                 (extra-inputs '())
                 (extra-envars '())
                 (builder "guix.gnu.org"))
 (let* ((version-major version-minor (if (pair? version)
                                       (car+cdr version)
                                       (values version #f))))
 (package
   (name (jname "jsoftware-j" release-type))
   (version (jversion->string version-major version-minor revision commit))
   (source
    (origin
      (method git-fetch)
      (uri (git-reference
            (url "https://github.com/jsoftware/jsource")
            (commit (or commit tag
                        (jinfo->git-tag version-major
                                        release-type
                                        version-minor)))))
      (sha256 (base32 hash))
      (file-name (git-file-name name version))
      (patches patches)))
   (build-system gnu-build-system)
   (native-inputs (list clang-toolchain))
   (inputs (cons* `("libedit" ,libedit)
                  `("libomp" ,libomp)
                  `("ijconsole" ,ijconsole)
                   ;; profilex.ijs overrides ~install and ~addons
                   ;; directories to reside under the user-writable ~user.
                   ;; This allows local-install of addons via pacman.  TODO:
                   ;; Guix-ify J addons as well.
                  `("profilex" ,(search-auxiliary-file
                                  "jsoftware/profilex.ijs"))
                  extra-inputs))
   (arguments
    `(#:modules (((ice-9 ftw) #:select (scandir))
                 ((ice-9 popen) #:select (open-pipe* close-pipe))
                 ((ice-9 regex) #:select (match:substring string-match))
                 ((ice-9 threads) #:select (parallel par-for-each))
                 ((srfi srfi-26) #:select (cut))
                 ((srfi srfi-1) #:select (fold))
                 ,@%gnu-build-system-modules)
      #:phases
      ;; Upstream's build system consists of ad-hoc scripts that build build up
      ;; (very complicated) environment variables to pass to make.  The basic
      ;; build process looks like this:
      ;;
      ;;   1) Copy jsrc/jversion-x.h to jsrc/jversion.h and edit values;
      ;;   2) Set jplatform and j64x environment variables;
      ;;   3) Run make2/build_jconsole.sh and make2/build_libj.sh;
      ;;
      ;; However, upstream expects users to run J directly from the source
      ;; directory; they do not supply a make `install' target.  Thus it takes
      ;; some massaging to install files in FHS-style directories.
      (modify-phases %standard-phases
        ;; In particular, we have to set up
        ;;
        ;;   1) jsrc/jversion.h as in a typical build;
        ;;   2) jlibrary/bin/profilex.ijs to point to writable directories;
        ;;   3) make2/build_*.sh to respect standard build conventions;
        ;;   4) jsrc/jconsole.c to fix libedit dlopen; and
        ;;   5) Hard coded references to addons directory.
        (replace 'configure
          (lambda* (#:key target inputs outputs #:allow-other-keys)
            (let* ((clang-toolchain (assoc-ref inputs "clang-toolchain"))
                   (clang (string-append clang-toolchain "/bin/clang"))
                   (libedit (assoc-ref inputs "libedit"))
                   (out (assoc-ref outputs "out")))
              ;; Set up build constants
              (copy-file "jsrc/jversion-x.h" "jsrc/jversion.h")
              (substitute* "jsrc/jversion.h"
                (("^#define jversion.*$")
                 (format #f "#define jversion ~s\n" ,version-major))
                (("^#define jtype.*$")
                 (format #f "#define jtype ~s\n"
                         ,(jrelease-string release-type version-minor)))
                (("^#define jbuilder.*$")
                 (format #f "#define jbuilder ~s\n" ,builder)))
              ;; Munge the build scripts into reason:
              ;; 1. Short-circuit the fragile compiler detection;
              ;; 2. Make sure to include our CFLAGS and LFLAGS; and
              ;; 3. Propagate script errors to top level.
              (for-each
               (lambda (file)
                 (with-directory-excursion "make2"
                   (substitute* file
                     ;; The `compiler' variable doesn't point to the actual
                     ;; compiler.  It is just a switch to tell the build
                     ;; scripts whether to use gcc- or clang-specific flags.
                     (("^compiler=.*$") "compiler=clang\n")
                     (("^LDFLAGS=\"" def) (string-append def "$LDFLAGS "))
                     (("^(common=\")(\\$USETHREAD.*)$" _ def rest)
                      (string-append def "$CFLAGS " rest))
                     (("^#!.*" shebang)
                      (string-append shebang "set -o errexit\n")))))
                 '("build_jconsole.sh" "build_libj.sh"))
              ;; The jconsole manually loads libedit with dlopen.  The path
              ;; must be absolute to correctly point to our input.
              (substitute* "jsrc/jconsole.c"
                (("libedit\\.so\\.[0-9]" so-file)
                 (format #f "~a/lib/~a" libedit so-file)))
              ;; The ~addons/dev directory supplies tentative J-script
              ;; definitions of new J engine functionality.  Since we point
              ;; ~addons under the ~user directory, we move it under ~system
              ;; instead, which sits as-is in the output.
              (with-directory-excursion "jsrc"
                (for-each
                  (lambda (file)
                    (substitute* file (("~addons/dev") "~system/dev")))
                  (scandir "."
                    (lambda (f) (eq? (stat:type (stat f)) 'regular)))))
              ;; Implementation of 9!:14 records build time which breaks build
              ;; reproducibility.  Note that upstream code depends on the exact
              ;; format of these strings, so we need to mimic the standard.
              (substitute* "jsrc/j.c"
                (("__DATE__") "\"Jan 01 1970\"")
                (("__TIME__") "\"00:00:00\""))
              ;; Upstream recommends using clang, with GCC support being
              ;; second-class, often resulting in build failures.
              (setenv "CC" clang))))

        ;; The build output depends primarily on the values of the `jplatform'
        ;; and `j64x' environment variables.  If the target is ARM, then
        ;; `jplatform' is "raspberry", otherwise it is `linux'.  In addition to
        ;; 32- and 64- bit versions, `j64x' controlls whether AVX or AVX2
        ;; variants of libj are built.
        ;;
        ;; However, build targets are not fine-grained enough to distinguish
        ;; between CPU features.  Thus we build and install all variants of
        ;; libj, expecting jconsole to be called with a wrapper script that
        ;; detects AVX features and loads the appropriate libj at runtime.
        (replace 'build
          (lambda _
            (setenv "USE_OPENMP" "1")
            (setenv "USE_THREAD" "1")
            (for-each (lambda (var-val) (apply setenv var-val))
                      (quote ,extra-envars))
            ;; The build scripts assume that PWD is make2.
            (with-directory-excursion "make2"
              (let* ((platform ,(if (target-arm?) "raspberry" "linux"))
                     (target-bit ,(if (target-64bit?) "64" "32"))
                     (run (lambda* (script #:key (avx ""))
                            (invoke "env"
                                    (string-append "jplatform=" platform)
                                    (string-append "j64x=j" target-bit avx)
                                    script))))
                (parallel
                  ;; Since jconsole doesn't depend on AVX features, we just
                  ;; build it once.
                  (run "./build_jconsole.sh")
                  (run "./build_libj.sh")
                  (if ,(target-64bit?)
                    (parallel
                      (run "./build_libj.sh" #:avx "avx")
                      (run "./build_libj.sh" #:avx "avx2"))))))))
        ;; The test suite is expected to be run as follows for each variant of
        ;; libj that we build:
        ;;
        ;;     $ echo 'RUN ddall' | jconsole test/tsu.ijs
        ;;
        ;; This requires a working jconsole with accessible jlibrary files.  We
        ;; simply place these all under test/bin.
        (replace 'check
          (lambda* (#:key tests? #:allow-other-keys)
            (when tests?
              (let ((platform ,(if (target-arm?) "raspberry" "linux")))
                (mkdir-p "test/bin")
                (for-each
                  (lambda (dir)
                    (let ((source (string-append "jlibrary/" dir))
                          (dest (string-append "test/bin/" dir)))
                    (begin
                      (mkdir-p dest)
                      (copy-recursively source dest))))
                  '("system" "tools" "addons"))
                ;; The jlibrary/dev directory only sometimes exists, but needs
                ;; to be copied into the ~system directory when it does.
                (for-each
                  (lambda (dev-dir)
                    (if (file-exists? dev-dir)
                      (copy-recursively dev-dir "test/bin/system/dev")))
                  '("jlibrary/dev" "jlibrary/addons/dev"))
                (par-for-each
                  (lambda (dir)
                    (let* ((bin (string-append "bin/" platform))
                           (jbit ,(if (target-64bit?) "j64" "j32"))
                           (jconsole (string-append bin "/" jbit
                                                    "/jconsole"))
                           (source (string-append bin "/" dir))
                           (dest (string-append "test/bin/" dir)))
                      (begin
                        (mkdir-p dest)
                        (copy-recursively source dest)
                        (install-file "jlibrary/bin/profile.ijs" dest)
                        (install-file jconsole dest)
                        (let* ((jconsole (string-append dest "/jconsole"))
                               (tests "test/tsu.ijs")
                               (port (open-pipe* OPEN_WRITE jconsole tests)))
                          (display "RUN ddall\n" port)
                          (when (not (zero? (status:exit-val
                                              (close-pipe port))))
                            (error "Some J build tests failed."))))))
                  (scandir (string-append "bin/" platform)
                           (negate (cut member <> '("." "..")))))
                #t))))
        ;; Now that everything is built, installation is fairly
        ;; straightforward, following FHS conventions.  The only quirk is that
        ;; we install jconsole under /libexec to make room for the wrapper
        ;; replacement under /bin.
        (replace 'install
          (lambda* (#:key outputs inputs #:allow-other-keys)
            (let* ((platform ,(if (target-arm?) "raspberry" "linux"))
                   (jbit ,(if (target-64bit?) "j64" "j32"))
                   (out (assoc-ref outputs "out"))
                   (bin (string-append out "/bin"))
                   (etc (string-append out "/etc/j"))
                   (lib (string-append out "/lib/j"))
                   (libexec (string-append out "/libexec/j"))
                   (share (string-append out "/share/j"))
                   (system (string-append share "/system"))
                   (dev (string-append system "/dev")))
              (mkdir-p bin)
              (copy-file (assoc-ref inputs "ijconsole")
                         (string-append bin "/ijconsole-" ,version-major))
              (mkdir-p lib)
              (for-each
                (lambda (jarch)
                  (let* ((jbin (string-join `("bin" ,platform ,jarch) "/"))
                         (javx-match (string-match "avx.*" jarch))
                         (javx (if (not javx-match) ""
                                 (match:substring javx-match)))
                         (sep (if javx-match "-" ""))
                         (source (string-append jbin "/libj.so"))
                         (dest (format #f "~a/libj~a~a.so" lib sep javx)))
                    (copy-file source dest)))
                (scandir (string-append "bin/" platform)
                         (negate (cut member <> '("." "..")))))
              (install-file (string-append "bin/" platform "/" jbit "/jconsole")
                            libexec)
              (copy-recursively "jlibrary/system" system)
              (for-each
                (lambda (source-dev)
                  (if (access? source-dev R_OK)
                    (copy-recursively source-dev dev)))
                '("jlibrary/dev" "jlibrary/addons/dev"))
              (install-file "jlibrary/bin/profile.ijs" etc)
              (copy-file (assoc-ref inputs "profilex")
                         (string-append etc "/profilex.ijs"))))))))
   (home-page "https://www.jsoftware.com/")
   (synopsis "Ascii-only, array programming language in the APL family")
   (description
    "J is a high-level, general-purpose programming language that is
particularly suited to the mathematical, statistical, and logical analysis of
data.  It is a powerful tool for developing algorithms and exploring problems
that are not already well understood.")
   (license license:gpl3+))))


(define-public jsoftware-j-901
  (make-j
    #:version '("901" . "f")
    #:hash "1776021m0j1aanzwg60by83n53pw7i6afd5wplfzczwk8bywax4p"
    #:patches (search-patches "jsoftware-j901-f-fixes.patch")))


(define j-build-configuration-with-sleef
  `(#:extra-inputs (("sleef" ,sleef))
    #:extra-envars (("USE_SLEEF_SRC" "0")
                    ("LDFLAGS" "-lsleef"))))

(define-public jsoftware-j-902
  (apply make-j
    (append j-build-configuration-with-sleef
      `(#:version ,'("902" . "b")
        #:hash "0j67vgikqflwjqacsdicasvyv1k54s2c8vjgwmf0ix7l41p4xqz0"))))


(define-public jsoftware-j-903
  (apply make-j
    (append j-build-configuration-with-sleef
      `(#:version ,'("903" . "a")
        #:tag "903-release-a"
        #:hash "1fcfl7q7c2vj4fmnqqc8c6hwgsjm20ff93v8xxfniasss1b2fmc4"))))


(define-public (jsoftware-ijconsole-symlink jpkg)
  "Provide bin/ijconsole symlink that points to pkg's bin/ijconsole-<jversion>"
  (package
    (name "jsoftware-ijconsole")
    (version (package-version jpkg))
    (source #f)
    (build-system trivial-build-system)
    (propagated-inputs `(("jpkg" ,jpkg)))
    (arguments
      `(#:modules ((guix build utils)
                   (srfi srfi-26))
        #:builder
        (begin
          (use-modules ((guix build utils) #:select (mkdir-p))
                       ((ice-9 regex) #:select (string-match))
                       ((ice-9 ftw) #:select (scandir))
                       ((srfi srfi-26) #:select (cut)))
          (let* ((out (assoc-ref %outputs "out"))
                 (jpkg (assoc-ref %build-inputs "jpkg"))
                 (ijconsole (car (scandir (string-append jpkg "/bin")
                                       (cut string-match "ijconsole-.*" <>))))
                 (source (string-append jpkg "/bin/" ijconsole))
                 (dest (string-append out "/bin/ijconsole")))
            (mkdir-p (dirname dest))
            (symlink source dest)))))
  (home-page (package-home-page jpkg))
  (synopsis "Provide `ijconsole' symlink to default interpreter version")
  (description
  "The interpreter provided by the J package has a filename like
ijconsole-<version>, which provides support for having multiple, concurrent
versions installed.  This package provides a version-agnostic `ijconsole'
symlink to interpreter version indicated and build time.")
  (license license:gpl3+)))

^ permalink raw reply	[flat|nested] 7+ messages in thread

* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
  2022-01-16  6:12 bug#53296: Intermittent segfaults when parsing (?) custom package from repo elaexuotee--- via Bug reports for GNU Guix
@ 2022-01-17 15:42 ` Ludovic Courtès
  2022-01-18  6:23   ` elaexuotee--- via Bug reports for GNU Guix
  0 siblings, 1 reply; 7+ messages in thread
From: Ludovic Courtès @ 2022-01-17 15:42 UTC (permalink / raw)
  To: elaexuotee; +Cc: 53296

Hi,

elaexuotee@wilsonb.com skribis:

> Recently working on a package, I have been encounting intermittent segfault
> during a build.
>
> The segfault seems to only occur when I have some error in my code that causes
> a crash, and the segfaults tend to cluster, appearing unexpectedly for a few
> build attempts, and then disappearing right as I think they are reproducible
> and try to grab an strace or something. Unfortunately, that's about the extent
> of information I have been able to gleen.
>
> The latest segfault happened with the attached package definition. Note, the
> offending code is at line 77, where I forgot to remove a docstring from a
> variable that used to be a procedure.

Here’s how I tried (and failed) to reproduce the segfault:

--8<---------------cut here---------------start------------->8---
$ find /tmp/test
/tmp/test
/tmp/test/gnu
/tmp/test/gnu/packages
/tmp/test/gnu/packages/jsoftware.scm

$ guix build -L /tmp/test jsoftware
/tmp/test/gnu/packages/jsoftware.scm:76:0: warning: source expression failed to match any pattern
guix build: error: jsoftware: nekonata pako

$ guix build -f /tmp/test/gnu/packages/jsoftware.scm 
/tmp/test/gnu/packages/jsoftware.scm:76:0: error: (define ijconsole "G-exp script that detects AVX/AVX2 support at runtime and executes jconsole\n  with the appropriate libj.so and profile.ijs." (with-imported-modules (quote ((guix cpu) (guix memoization) (guix profiling) (guix sets) (srfi srfi-26))) (program-file "ijconsole" (gexp (begin (use-modules ((guix cpu) #:select (cpu-flags current-cpu)) ((guix sets) #:select (set-contains?)) ((srfi srfi-26) #:select (cute))) (define %basedir (dirname (dirname (current-filename)))) (let* ((jconsole (string-append %basedir "/libexec/j/jconsole")) (cpu-has-flag? (cute set-contains? (cpu-flags (current-cpu)) <>)) (libj (format #f "~a/lib/j/libj~a.so" %basedir (cond ((cpu-has-flag? "avx2") "-avx2") ((cpu-has-flag? "avx") "-avx") (else "")))) (jprofile (string-append %basedir "/etc/j/profile.ijs"))) (apply execl jconsole "ijconsole" "-lib" libj "-jprofile" jprofile (cdr (command-line))))))))): source expression failed to match any pattern

$ guix describe
Generacio 201   Jan 12 2022 23:15:13    (nuna)
  guix 0052c3b
    repository URL: https://git.savannah.gnu.org/git/guix.git
    branch: master
    commit: 0052c3b0458fba32920a1cfb48b8311429f0d6b5
--8<---------------cut here---------------end--------------->8---

Is there another way to reproduce it?

Is the ‘LD_LIBRARY_PATH’ variable set on your system?

Thanks,
Ludo’.




^ permalink raw reply	[flat|nested] 7+ messages in thread

* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
  2022-01-17 15:42 ` Ludovic Courtès
@ 2022-01-18  6:23   ` elaexuotee--- via Bug reports for GNU Guix
  2022-01-18 11:28     ` Ludovic Courtès
  0 siblings, 1 reply; 7+ messages in thread
From: elaexuotee--- via Bug reports for GNU Guix @ 2022-01-18  6:23 UTC (permalink / raw)
  To: Ludovic Courtès; +Cc: 53296

Good day. Thanks for looking into this.

Ludovic Courtès <ludo@gnu.org> wrote:
> Here’s how I tried (and failed) to reproduce the segfault:
> 
> --8<---------------cut here---------------start------------->8---
> $ find /tmp/test
> /tmp/test
> /tmp/test/gnu
> /tmp/test/gnu/packages
> /tmp/test/gnu/packages/jsoftware.scm
> 
> $ guix build -L /tmp/test jsoftware
> /tmp/test/gnu/packages/jsoftware.scm:76:0: warning: source expression failed to match any pattern
> guix build: error: jsoftware: nekonata pako

Is that an Esperanto locale?!

> $ guix build -f /tmp/test/gnu/packages/jsoftware.scm 
> /tmp/test/gnu/packages/jsoftware.scm:76:0: error: (define ijconsole "G-exp script that detects AVX/AVX2 support at runtime and executes jconsole\n  with the appropriate libj.so and profile.ijs." (with-imported-modules (quote ((guix cpu) (guix memoization) (guix profiling) (guix sets) (srfi srfi-26))) (program-file "ijconsole" (gexp (begin (use-modules ((guix cpu) #:select (cpu-flags current-cpu)) ((guix sets) #:select (set-contains?)) ((srfi srfi-26) #:select (cute))) (define %basedir (dirname (dirname (current-filename)))) (let* ((jconsole (string-append %basedir "/libexec/j/jconsole")) (cpu-has-flag? (cute set-contains? (cpu-flags (current-cpu)) <>)) (libj (format #f "~a/lib/j/libj~a.so" %basedir (cond ((cpu-has-flag? "avx2") "-avx2") ((cpu-has-flag? "avx") "-avx") (else "")))) (jprofile (string-append %basedir "/etc/j/profile.ijs"))) (apply execl jconsole "ijconsole" "-lib" libj "-jprofile" jprofile (cdr (command-line))))))))): source expression failed to match any pattern
>
> $ guix describe
> Generacio 201   Jan 12 2022 23:15:13    (nuna)
>   guix 0052c3b
>     repository URL: https://git.savannah.gnu.org/git/guix.git
>     branch: master
>     commit: 0052c3b0458fba32920a1cfb48b8311429f0d6b5
> --8<---------------cut here---------------end--------------->8---
> 
> Is there another way to reproduce it?

The segfault is pretty inconsistent on my end. Unfortunately, the only
"reliable" way to produce I know at the moment is to badger the build command
repeatedly and get (un)lucky.

> Is the ‘LD_LIBRARY_PATH’ variable set on your system?

Good thought. But no; I am seeing the segfaults in a --pure shell.




^ permalink raw reply	[flat|nested] 7+ messages in thread

* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
  2022-01-18  6:23   ` elaexuotee--- via Bug reports for GNU Guix
@ 2022-01-18 11:28     ` Ludovic Courtès
  2022-03-08  8:46       ` Ludovic Courtès
  0 siblings, 1 reply; 7+ messages in thread
From: Ludovic Courtès @ 2022-01-18 11:28 UTC (permalink / raw)
  To: elaexuotee; +Cc: 53296

Hi,

elaexuotee@wilsonb.com skribis:

> Ludovic Courtès <ludo@gnu.org> wrote:
>> Here’s how I tried (and failed) to reproduce the segfault:
>> 
>> --8<---------------cut here---------------start------------->8---
>> $ find /tmp/test
>> /tmp/test
>> /tmp/test/gnu
>> /tmp/test/gnu/packages
>> /tmp/test/gnu/packages/jsoftware.scm
>> 
>> $ guix build -L /tmp/test jsoftware
>> /tmp/test/gnu/packages/jsoftware.scm:76:0: warning: source expression failed to match any pattern
>> guix build: error: jsoftware: nekonata pako
>
> Is that an Esperanto locale?!

Jes!

> The segfault is pretty inconsistent on my end. Unfortunately, the only
> "reliable" way to produce I know at the moment is to badger the build command
> repeatedly and get (un)lucky.

OK, please let us know when you have a reliable reproducer, or GDB
backtraces, things like that.

Thanks,
Ludo’.




^ permalink raw reply	[flat|nested] 7+ messages in thread

* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
  2022-01-18 11:28     ` Ludovic Courtès
@ 2022-03-08  8:46       ` Ludovic Courtès
  2023-02-12 13:07         ` elaexuotee--- via Bug reports for GNU Guix
  0 siblings, 1 reply; 7+ messages in thread
From: Ludovic Courtès @ 2022-03-08  8:46 UTC (permalink / raw)
  To: elaexuotee; +Cc: 53296

Hi,

Any update on this one?

  https://issues.guix.gnu.org/53296

If there’s no reliable way to reproduce it, I’ll close the bug soon.

Thanks,
Ludo’.

Ludovic Courtès <ludo@gnu.org> skribis:

> Hi,
>
> elaexuotee@wilsonb.com skribis:
>
>> Ludovic Courtès <ludo@gnu.org> wrote:
>>> Here’s how I tried (and failed) to reproduce the segfault:
>>> 
>>> --8<---------------cut here---------------start------------->8---
>>> $ find /tmp/test
>>> /tmp/test
>>> /tmp/test/gnu
>>> /tmp/test/gnu/packages
>>> /tmp/test/gnu/packages/jsoftware.scm
>>> 
>>> $ guix build -L /tmp/test jsoftware
>>> /tmp/test/gnu/packages/jsoftware.scm:76:0: warning: source expression failed to match any pattern
>>> guix build: error: jsoftware: nekonata pako
>>
>> Is that an Esperanto locale?!
>
> Jes!
>
>> The segfault is pretty inconsistent on my end. Unfortunately, the only
>> "reliable" way to produce I know at the moment is to badger the build command
>> repeatedly and get (un)lucky.
>
> OK, please let us know when you have a reliable reproducer, or GDB
> backtraces, things like that.
>
> Thanks,
> Ludo’.




^ permalink raw reply	[flat|nested] 7+ messages in thread

* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
  2022-03-08  8:46       ` Ludovic Courtès
@ 2023-02-12 13:07         ` elaexuotee--- via Bug reports for GNU Guix
  2023-02-14 12:04           ` Simon Tournier
  0 siblings, 1 reply; 7+ messages in thread
From: elaexuotee--- via Bug reports for GNU Guix @ 2023-02-12 13:07 UTC (permalink / raw)
  To: Ludovic Courtès; +Cc: 53296

Nothing like ACKing a year later!

This was almost certainly a local hardware issue. The machine this happened on
ended up showing progressively more mysterious behaviour that could not be
reproduced on other machines.

Feel free to close!

Ludovic Courtès <ludo@gnu.org> wrote:
> Hi,
> 
> Any update on this one?
> 
>   https://issues.guix.gnu.org/53296
> 
> If there’s no reliable way to reproduce it, I’ll close the bug soon.
> 
> Thanks,
> Ludo’.
> 
> Ludovic Courtès <ludo@gnu.org> skribis:
> 
> > Hi,
> >
> > elaexuotee@wilsonb.com skribis:
> >
> >> Ludovic Courtès <ludo@gnu.org> wrote:
> >>> Here’s how I tried (and failed) to reproduce the segfault:
> >>> 
> >>> --8<---------------cut here---------------start------------->8---
> >>> $ find /tmp/test
> >>> /tmp/test
> >>> /tmp/test/gnu
> >>> /tmp/test/gnu/packages
> >>> /tmp/test/gnu/packages/jsoftware.scm
> >>> 
> >>> $ guix build -L /tmp/test jsoftware
> >>> /tmp/test/gnu/packages/jsoftware.scm:76:0: warning: source expression failed to match any pattern
> >>> guix build: error: jsoftware: nekonata pako
> >>
> >> Is that an Esperanto locale?!
> >
> > Jes!
> >
> >> The segfault is pretty inconsistent on my end. Unfortunately, the only
> >> "reliable" way to produce I know at the moment is to badger the build command
> >> repeatedly and get (un)lucky.
> >
> > OK, please let us know when you have a reliable reproducer, or GDB
> > backtraces, things like that.
> >
> > Thanks,
> > Ludo’.






^ permalink raw reply	[flat|nested] 7+ messages in thread

* bug#53296: Intermittent segfaults when parsing (?) custom package from repo.
  2023-02-12 13:07         ` elaexuotee--- via Bug reports for GNU Guix
@ 2023-02-14 12:04           ` Simon Tournier
  0 siblings, 0 replies; 7+ messages in thread
From: Simon Tournier @ 2023-02-14 12:04 UTC (permalink / raw)
  To: elaexuotee; +Cc: 53296-done, Ludovic Courtès

Hi,

>> If there’s no reliable way to reproduce it, I’ll close the bug soon.
>
> This was almost certainly a local hardware issue. The machine this happened on
> ended up showing progressively more mysterious behaviour that could not be
> reproduced on other machines.
>
> Feel free to close!

So closing!

Thanks,
simon




^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2023-02-14 12:26 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2022-01-16  6:12 bug#53296: Intermittent segfaults when parsing (?) custom package from repo elaexuotee--- via Bug reports for GNU Guix
2022-01-17 15:42 ` Ludovic Courtès
2022-01-18  6:23   ` elaexuotee--- via Bug reports for GNU Guix
2022-01-18 11:28     ` Ludovic Courtès
2022-03-08  8:46       ` Ludovic Courtès
2023-02-12 13:07         ` elaexuotee--- via Bug reports for GNU Guix
2023-02-14 12:04           ` Simon Tournier

Code repositories for project(s) associated with this external index

	https://git.savannah.gnu.org/cgit/guix.git

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.