1748 replies to this topic

[Erik Slagter]

I guess it means the suppression of the typical "vibration" of thin horizontal lines when scaling interlaced video, due to the unavoidable bob/weaving before the scaling.

[Persian Prince]

i think it won't harm if you add the support

 

any future drivers with the related proc files could support it

[SpaceRat]

Fix for OpenSSL (CVE-2014-0160) Heartbleed [SECURITY-ALERT!]

--- openpli-oe-core/openembedded-core/meta/recipes-connectivity/openssl/openssl_1.0.1e.old	2014-04-12 17:34:17.169024482 +0200
+++ openpli-oe-core/openembedded-core/meta/recipes-connectivity/openssl/openssl_1.0.1e.bb	2014-04-12 17:33:49.565024089 +0200
@@ -4,7 +4,7 @@
 # if they are available.
 DEPENDS += "ocf-linux"
 
-CFLAG += "-DHAVE_CRYPTODEV -DUSE_CRYPTODEV_DIGESTS"
+CFLAG += "-DHAVE_CRYPTODEV -DUSE_CRYPTODEV_DIGESTS -DOPENSSL_NO_HEARTBEATS"
 
 PR = "${INC_PR}.0"
 

After applying the OpenSSL fix, you can bumb OpenVPN to 2.3.3 (As I do not know which .bb is used, here without path):

--- openvpn_2.3.2.bb    2014-04-10 10:21:33.952676200 +0200
+++ openvpn_2.3.3.bb    2014-04-10 11:18:24.468938800 +0200
@@ -10,8 +10,8 @@
 SRC_URI = "http://swupdate.openvpn.org/community/releases/openvpn-${PV}.tar.gz \
            file://openvpn"

-SRC_URI[md5sum] = "06e5f93dbf13f2c19647ca15ffc23ac1"
-SRC_URI[sha256sum] = "20bda3f9debb9a52db262aecddfa4e814050a9404a9106136b7e3b6f7ef36ffc"
+SRC_URI[md5sum] = "5c66ea3143ac884a3075521bd74ede06"
+SRC_URI[sha256sum] = "f025d14631105a66e501ca897830cd4d26a1438530cd9174dc6169536ae4b113"

 CFLAGS += "-fno-inline"
 

[christophecvr]

I noticed that with latest build for dm8000, the gst-dreambox-mediasink is updated. The dts-passtrough-downpatch-for-ac3 not needed anymore. It is already adapted into the basic source.

The patch fails ass well during build wich is logic.

 

Here the git patch which removes this dts-pastrough patch for gst-dreambox-dvbmediasink  build.

 

Tested on dm8000 and all passtrough ac3 works perfect (even ac3+ by me since I patched enigma2 in my own build).

 

git patch against master off 08:30 Hour this morning.

 

 

Attached Files

•  Dts-PasstroughPatch-Removal.patch   4.2KB   21 downloads

[littlesat]

@spacerat,

Can you offer us a patch we can directly commit using git -am

[athoik]

I noticed that with latest build for dm8000, the gst-dreambox-mediasink is updated. The dts-passtrough-downpatch-for-ac3 not needed anymore.

DTS != AC3

Are you sure you can get DTS sound without a patch?

[christophecvr]

I noticed that with latest build for dm8000, the gst-dreambox-mediasink is updated. The dts-passtrough-downpatch-for-ac3 not needed anymore.

DTS != AC3

Are you sure you can get DTS sound without a patch?

Is ac3+ DTS ?

 

Then Yes (on dm off course with enigma2 patch ass well which is not implemented in openpli4 basic)

 

Which channels are dts or have dts stream so I can test I indeed did only acp and ac3+

Edited by christophecvr, 13 April 2014 - 11:09.

[ims]

And what line "SRC_URI += "file://dm-ac3Plusbypass.patch" in enigma.bb ?

[christophecvr]

And what line "SRC_URI += "file://dm-ac3Plusbypass.patch" in enigma.bb ?

gues You mean this ac3+ patch for dreamboxes included here.

 

That is still needed has nothing to do with gst-dreambox-dvbmediasink but it's the dm drivers self which uses bypass 7 insstead off 0x22 like drivers for other stb's.

Attached Files

•  cvrMydmac3.patch   2.02KB   4 downloads

[ims]

ok, but then it should not be added into patch for DTS...Removal...

[christophecvr]

ok, but then it should not be added into patch for DTS...Removal...

Do not follow what You mean. here ?

[christophecvr]

I noticed that with latest build for dm8000, the gst-dreambox-mediasink is updated. The dts-passtrough-downpatch-for-ac3 not needed anymore.

DTS != AC3

Are you sure you can get DTS sound without a patch?

Ard das erste = with dolby digital 2.0 and works fine without that 001-passtrough patch

[christophecvr]

Also arte HD works in Originalton Dolby Digital  So indeed that patch is not needed anymore.

[christophecvr]

ok, but then it should not be added into patch for DTS...Removal...

Do not follow what You mean. here ?

Ohh shit yes I see now what You mean (I forgot that I commited that personal change before thats why it creeped unwanted in the patch) sorry sorry.

 

Hereby the corrected one

Attached Files

•  Dts-PasstroughPatch-Removal.patch   3.73KB   3 downloads

Edited by christophecvr, 13 April 2014 - 13:00.

[SpaceRat]

@spacerat,
Can you offer us a patch we can directly commit using git -am

Just merge openembedded-core, they have updated to OpenSSL 1.0.1g (rather than adding just another patch to 1.0.1e):
http://git.openembed...ctivity/openssl

[christophecvr]

@spacerat,
Can you offer us a patch we can directly commit using git -am

Just merge openembedded-core, they have updated to OpenSSL 1.0.1g (rather than adding just another patch to 1.0.1e):
http://git.openembed...ctivity/openssl

I seems that the git was already patched with that 1.0.1e cve patch

 

openssl: backport fix for CVE-2014-0160

Just upgrading OE-core was causing too much trouble for now, so
instead, patch the vulnerability that really matters. Also bump
the PR.

See also:
http://cgit.openembe...84dd90177379242

[SpaceRat]

openssl: backport fix for CVE-2014-0160

Head <> wall.

If I was an author, I would add this to my license:

"x.1 The name of this software is $NAME.
x.2 Its name together with a version number assigned by its original author(s) denotes a specific version of this product.
x.3 Each version number equals exactly one (1) source code set on the author(s) code repository.
x.4 You are not allowed to be suggestive to be delivering a specific version of this program unless you are using its original code as pointed out under x.3.
...
x.y Programs based on a multitude of versions of this software and/or containing hacks ressembling the behaviour of newer versions in older versions ("backports") of it MUST NOT be named $NAME and MUST NOT be named in a way suggesting it was software $NAME.
x.z Hacked versions of $NAME MUST NOT contain links to original support addresses, neither eMail nor any other form of URI or hint at original support places."


@christophecvr: YOU know why I'm that pissed off by fake versions.
Obviously the idiots at Debian don't know everything but everything better ...
So they should write their own OS and programs, rather than hacking existing code, faking versions.

This is the mess they left us with:

All OpenSSL versions 1.0.1 to 1.0.1e are vulnerable unless the heartbeat functionality has been disabled, which can easily be detected by just checking for heartbeats.
They are also not vulnerable if they have been uselessly been patched rather than updated, but this can only be guessed or by trying to exploit, as patched versions have heartbeats.

All other versions are not affected UNLESS they received heartbeat functionality through backports from versions OpenSSL 1.0.1 to 1.0.1e UNLESS a patched version was used for the backport ...

Gnarf, argls, vomit.

Oh and yes:
All versions of vlc 2.1.3 can handle streams transcoded by Vu+ Solo² or Duo² UNLESS a set of unfucked idiots with ash-tray sized glasses at Debian decided to ship a version of libav from 1965 with your OS ...

[WanWizard]

@SpaceRat

 

We're looking into the oe-core merge at the moment, which seams like a better option than disabling the heartbeat or applying the patch.

[christophecvr]

@SpaceRat,

ter info

 

I do have nothing to do with this openssl: backport fix for CVE-2014-0160 fix.

 

Just repported that someone already put it in te git.

[Erik Slagter]

All versions of vlc 2.1.3 can handle streams transcoded by Vu+ Solo² or Duo² UNLESS a set of unfucked idiots with ash-tray sized glasses at Debian decided to ship a version of libav from 1965 with your OS ...

This has nothing to do with debian. All distributions are guilty of crippling libav, either by disabling codecs or by issueing stone aged version ("we only ship released software"). We all know that, there is no need for such language.