There is a regression in vcut that caused a segmentation fault every time I run this program.
It was introduced by commit f7cb48a75de43af793b0399382faec39fcb9b495 (Flush stream after the third header packet, and delay the writing of … )
In submit_headers_to_stream(), vs->headers is only 3 structures long (as defined in vcut.h) so accessing vs->headers[3] is certainly a mistake. Moreover, in this function, i is always smaller then 4 so the comparision if(i < 4) makes no sense. The Author probably intended to write if (i<3), which fits with the general modifications of submit_hedears_to_stream in commit f7cb48a75de43af793b0399382faec39fcb9b495 .
I am running Ubuntu 16.04 i386, maybe this bug does not cause segmentation faults on 64-bit architectures.
There is a regression in vcut that caused a segmentation fault every time I run this program. It was introduced by commit f7cb48a75de43af793b0399382faec39fcb9b495 (Flush stream after the third header packet, and delay the writing of … )
In
submit_headers_to_stream(),vs->headersis only 3 structures long (as defined invcut.h) so accessingvs->headers[3]is certainly a mistake. Moreover, in this function, i is always smaller then 4 so the comparisionif(i < 4)makes no sense. The Author probably intended to writeif (i<3), which fits with the general modifications ofsubmit_hedears_to_streamin commit f7cb48a75de43af793b0399382faec39fcb9b495 .I am running Ubuntu 16.04 i386, maybe this bug does not cause segmentation faults on 64-bit architectures.