5 ms·
Ghostscript zero-day allows full server compromises
- jbverschoor 5y agoDuplicate of https://news.ycombinator.com/item?id=28446810 https://news.ycombinator.com/item?id=28446810
- jessaustin 5y agoThere is no need to point out previous empty discussions.
- lioeters 5y ago> Ghostscript is..typically included with image conversion and file upload processing toolkits, such as the popular ImageMagick. > The proof-of-concept code..[allows] an attacker to upload a malformed SVG file that escapes the image processing pipeline and runs malicious code on the underlying operating system. Python code demonstrating the exploit: https://github.com/duc-nt/RCE-0-day-for-GhostScript-9.50/blob/main/IM-RCE-via-GhostScript-9.5.py https://github.com/duc-nt/RCE-0-day-for-GhostScript-9.50/blo...
- tyingq 5y agoThe article mentions a similar exploit in 2018. There was also one in 2009: https://www.cvedetails.com/cve/CVE-2009-0196/ https://www.cvedetails.com/cve/CVE-2009-0196/
- mavhc 5y agoRunning complex parsers not in a sandbox? Did they not learn from all of history?
- Caligatio 5y agoGhostscript was initially released 1988 (per Wikipedia) so it definitely is more in the history making rather than learning from history category.
- pabs3 5y ago> However, Artifex, the company behind the Ghostscript project, told The Record the vulnerability had not been reported to the company via its vulnerability disclosure process. > The company said they are “becoming increasingly frustrated with security researchers who fail to ethically disclose potentially serious security vulnerabilities” and that engineers are currently working on a patch they hope to have out by the end of the week. Hmmm.
- krageon 5y agoThey offer $2000 bounties max. It's no wonder the researcher didn't bother disclosing to them (assuming this is even true) and instead disclosed to downstream companies, that is not a reasonable payment commensurate to the effort and criticality of the finding. Folks are not obligated to donate their time.
- tommiegannert 5y ago>> While Nguyen released the public exploit for this bug, he is not the one who discovered the vulnerability. >> >> The person who did is Wunderfund CTO and founder Emil Lerner, who found the bug last year and used it to obtain bug bounties from companies like Airbnb, Dropbox, and Yandex. > > Folks are not obligated to donate their time. When it comes to open-source, I think this is a weak argument. While I consider your statement true, it's a "glass half-empty" way of looking at working together to build useful software. Releasing an exploit before a fix is applied also negates work other people have put into the project. It's surprising to me that neither Lerner, nor any of the downstream companies, upstreamed the fix in a year's time. If true that none tried, as you point out. > Details about the vulnerability leaked into the public domain last month after Lerner held a talk at the ZeroNight X security conference about the current attack vector posed by server-side image conversion tools and used the Ghostscript zero-day as an example.
- pabs3 5y agoWhile I can understand trying to get larger payouts and I still think what the researcher did is unethical, I'm even more disappointed with the security process of those downstream companies, they have to be pretty incompetent to not immediately burn the vulnerability by reporting it upstream. Unless of course they were planning on exploiting their competitors using it, which is even worse. Probably more likely is employees are forbidden from contributing back to open source, which means they each have incompetent legal departments.
- pabs3 5y agoI guess this is the patch for the issue: https://git.ghostscript.com/?p=ghostpdl.git;a=commitdiff;h=a9bd3dec9fde https://git.ghostscript.com/?p=ghostpdl.git;a=commitdiff;h=a...