Feed entropy into random pool
The haveged daemon feeds the linux entropy pool with random
numbers generated from hidden processor state.
For more informations see http://www.issihosts.com/haveged/
- Developed at security
- Sources inherited from project openSUSE:Factory
-
4
derived packages
- Download package
-
Checkout Package
osc -A https://api.opensuse.org checkout openSUSE:Factory:PowerPC/haveged && cd $_
- Create Badge
Refresh
Refresh
Source Files
Filename | Size | Changed |
---|---|---|
90-haveged.rules | 0000000107 107 Bytes | |
haveged-1.9.1.tar.gz | 0000474845 464 KB | |
haveged-conditional-enttest.patch | 0000001099 1.07 KB | |
haveged.changes | 0000012980 12.7 KB | |
haveged.service | 0000000301 301 Bytes | |
haveged.spec | 0000004149 4.05 KB | |
ppc64le.patch | 0000000407 407 Bytes |
Revision 39 (latest revision is 66)
Stephan Kulow (coolo)
accepted
request 233562
from
Tomáš Chvátal (scarabeus_factory)
(revision 39)
- Disable randomly failing ENT test bnc#876674 - Added patches: * haveged-conditional-enttest.patch - Threads seem not to fix the random testsuite fail. - Fix "stop" of service, the daemon in foreground actually exits with a weird exit code. workaround this misbehaviour in the service file. - revert to non-forking service, "forking" services are racy and provided by systemd purely for compatibility with old daemons. see daemon(7) for the rationale. Only "simple" (default) and "notify" services must be used if possible. - Also, there is no haveged.conf around, neither such file is parsed by the daemon at all. - Start the daemon as soon as the random device is available with the help of udev, as starting services while starved of entropy is no good.
Comments 1
Haveged is no longer necessary since around kernel 5.6.. at least we need to drop the udev rules, never activate it as default and drop all the initrd parts.