You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In both scheduler/cupsfilter.c and scheduler/job.c, there's a rip_max_cache variable which gets included in the environment array for subprocesses. However, it's never initialised, so what actually ends up in the environment is some random junk from the stack.
This causes a test failure for me with current master (on amd64 Linux with GCC 11.2), because error.log ends up containing binary junk and the test suite's calls to grep report that it's a binary file.
error.log:
D [12/Jan/2022:11:50:51.840244 +0000] [Job 66] envp[23]="CHARSET=utf-8"
D [12/Jan/2022:11:50:51.840245 +0000] [Job 66] envp[24]="LANG=en.UTF-8"
D [12/Jan/2022:11:50:51.840246 +0000] [Job 66] envp[25]="PPD=/tmp/cups-garstow/ppd/Test1.ppd"
D [12/Jan/2022:11:50:51.840247 +0000] [Job 66] envp[26]="\006ïÔL�"
D [12/Jan/2022:11:50:51.840248 +0000] [Job 66] envp[27]="CONTENT_TYPE=image/jpeg"
D [12/Jan/2022:11:50:51.840249 +0000] [Job 66] envp[28]="DEVICE_URI=file:/tmp/Test1"
Test suite output:
Test Summary
PASS: cupsd exited with no errors.
PASS: All job control files purged.
PASS: Printer 'Test1' correctly produced 131 page(s).
PASS: Printer 'Test2' correctly produced 23 page(s).
PASS: Printer 'Test3' correctly produced 2 page(s).
PASS: 176 requests logged.
PASS: CUPS-Get-Default not logged.
PASS: 0 emergency messages.
PASS: 0 alert messages.
PASS: 0 critical messages.
PASS: 33 error messages.
grep: /tmp/cups-garstow/log/error_log: binary file matches
FAIL: 4 warning messages, expected 14.
W [12/Jan/2022:11:46:13.686312 +0000] No JobPrivateAccess defined in policy default - using defaults.
W [12/Jan/2022:11:46:13.686315 +0000] No JobPrivateValues defined in policy default - using defaults.
W [12/Jan/2022:11:46:13.686316 +0000] No SubscriptionPrivateAccess defined in policy default - using defaults.
W [12/Jan/2022:11:46:13.686317 +0000] No SubscriptionPrivateValues defined in policy default - using defaults.
grep: /tmp/cups-garstow/log/error_log: binary file matches
grep: /tmp/cups-garstow/log/error_log: binary file matches
PASS: 0 notice messages.
grep: /tmp/cups-garstow/log/error_log: binary file matches
PASS: 137 info messages.
grep: /tmp/cups-garstow/log/error_log: binary file matches
PASS: 4243 debug messages.
grep: /tmp/cups-garstow/log/error_log: binary file matches
PASS: 6135 debug2 messages.
grep: /tmp/cups-garstow/log/error_log: binary file matches
Removing the rip_max_cache code entirely from those two files fixes this for me, but I'm not sure if that's the right fix or if it's intended to contain something...
The text was updated successfully, but these errors were encountered:
In both
scheduler/cupsfilter.c
andscheduler/job.c
, there's arip_max_cache
variable which gets included in the environment array for subprocesses. However, it's never initialised, so what actually ends up in the environment is some random junk from the stack.This causes a test failure for me with current master (on amd64 Linux with GCC 11.2), because
error.log
ends up containing binary junk and the test suite's calls to grep report that it's a binary file.error.log
:Test suite output:
Removing the
rip_max_cache
code entirely from those two files fixes this for me, but I'm not sure if that's the right fix or if it's intended to contain something...The text was updated successfully, but these errors were encountered: