Re: [poppler] Lots of fuzzing issues in cairo related code

2020-12-09 Thread Albert Astals Cid 
El dimecres, 9 de desembre de 2020, a les 19:04:37 CET, Germán Poo-Caamaño va 
escriure:
> On Wed, 2020-12-09 at 18:53 +0100, Albert Astals Cid wrote:
> > We recently enabled oss-fuzz on CairoOutputDev and that has found
> > lots of issues, is there anyone that wants to look at them (will need
> > a google-enabled account) before they are made public by the oss-fuzz
> > timeout?
> 
> What is the oss-fuzz timeout?

90 days.

Cheers,
  Albert

> 
> I would like to take a look at it, even though I lack of spare time, my
> backlog keeps increasing, and my cairo knowledge is limited at best.
> 
> 




___
poppler mailing list
poppler@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/poppler

Re: [poppler] Lots of fuzzing issues in cairo related code

2020-12-09 Thread Germán Poo-Caamaño 
On Wed, 2020-12-09 at 18:53 +0100, Albert Astals Cid wrote:
> We recently enabled oss-fuzz on CairoOutputDev and that has found
> lots of issues, is there anyone that wants to look at them (will need
> a google-enabled account) before they are made public by the oss-fuzz
> timeout?

What is the oss-fuzz timeout?

I would like to take a look at it, even though I lack of spare time, my
backlog keeps increasing, and my cairo knowledge is limited at best.

-- 
Germán Poo-Caamaño
https://calcifer.org




signature.asc
Description: This is a digitally signed message part
___
poppler mailing list
poppler@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/poppler

[poppler] Lots of fuzzing issues in cairo related code

2020-12-09 Thread Albert Astals Cid 
We recently enabled oss-fuzz on CairoOutputDev and that has found lots of 
issues, is there anyone that wants to look at them (will need a google-enabled 
account) before they are made public by the oss-fuzz timeout?

Cheers,
  Albert


___
poppler mailing list
poppler@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/poppler

[poppler] poppler/Gfx.cc

2020-12-09 Thread GitLab Mirror 
 poppler/Gfx.cc |7 ---
 1 file changed, 4 insertions(+), 3 deletions(-)

New commits:
commit 4601e1d533a9c47e304b75435a104f5096a2b9a4
Author: Philipp Knechtges 
Date:   Tue Dec 8 23:06:43 2020 +0100

Gfx: specifically use DeviceGray instead of DefaultGray for softmasks

diff --git a/poppler/Gfx.cc b/poppler/Gfx.cc
index b8bf47ab..746703ab 100644
--- a/poppler/Gfx.cc
+++ b/poppler/Gfx.cc
@@ -4392,11 +4392,12 @@ void Gfx::doImage(Object *ref, Stream *str, bool 
inlineImg)
 obj1 = std::move(obj2);
 }
 }
-maskColorSpace = GfxColorSpace::parse(nullptr, , out, state);
-if (!maskColorSpace || maskColorSpace->getMode() != csDeviceGray) {
-delete maskColorSpace;
+// Here, we parse manually instead of using GfxColorSpace::parse,
+// since we explicitly need DeviceGray and not some DefaultGray 
color space
+if (!obj1.isName("DeviceGray") && !obj1.isName("G")) {
 goto err1;
 }
+maskColorSpace = new GfxDeviceGrayColorSpace();
 obj1 = maskDict->lookup("Decode");
 if (obj1.isNull()) {
 obj1 = maskDict->lookup("D");
___
poppler mailing list
poppler@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/poppler