Bug#1013192: linux-image-5.10.0-15-amd64: ridiculously small entropy pool

2022-06-18 Thread Diederik de Haas
Control: severity -1 normal
Control: merge -1 1012835

On Saturday, 18 June 2022 21:52:41 CEST Thorsten Glaser wrote:
> Version: 5.10.120-1
> Severity: serious
> Tags: security

There has been a HUGE changeset applied between 5.10.118 and 5.10.119 and while
not entirely certain, I'm quite confident that this change was intentional. 
That and the severity of the bug I'm merging it with has severity normal, is
the reason I'm downgrading the severity to normal.
I'll leave it up to the maintainer to adjust it if needed.

https://lore.kernel.org/all/20220317232804.931702-1-ja...@zx2c4.com/ is 
probably the closest match of the 'cause' of these changes, but there's a good
chance that several patch sets were involved.

Here are some more 'random' threads I have found:
https://lore.kernel.org/all/20211221175047.341782-1-ja...@zx2c4.com/
https://lore.kernel.org/all/20220201161342.154666-1-ja...@zx2c4.com/

And as already mentioned in the bug I'm merging it with, you're not the only
one who noticed:
https://forum.openwrt.org/t/low-entropy-22-03-snapshot-change-in-kernel-entropy-pool-logic/129573

signature.asc
Description: This is a digitally signed message part.


Bug#1013192: linux-image-5.10.0-15-amd64: ridiculously small entropy pool

2022-06-18 Thread Thorsten Glaser
Bastian Blank dixit:

>The pool size for an RPNG is only the size of the state, nothing else.

Yes, and that is the problem. It was small before, it’s ridiculous now.

>might not have had any value before anyway.  You just need to reseed on
>a regular interval.

Ugh. I recall reading something about this on LWN, but I thought I
had time until bookworm to invent something to deal with this…

bye,
//mirabilos
-- 
(gnutls can also be used, but if you are compiling lynx for your own use,
there is no reason to consider using that package)
-- Thomas E. Dickey on the Lynx mailing list, about OpenSSL



Bug#1013192: linux-image-5.10.0-15-amd64: ridiculously small entropy pool

2022-06-18 Thread Diederik de Haas
On Saturday, 18 June 2022 22:47:01 CEST Diederik de Haas wrote:
> Here are some more 'random' threads I have found:

And this seems like an entire document explaining it:
https://www.zx2c4.com/projects/linux-rng-5.17-5.18/

signature.asc
Description: This is a digitally signed message part.


Bug#1013192: linux-image-5.10.0-15-amd64: ridiculously small entropy pool

2022-06-18 Thread Bastian Blank
Control: severity -1 normal
Control: tags -1 wontfix

On Sat, Jun 18, 2022 at 09:52:41PM +0200, Thorsten Glaser wrote:
> /proc/sys/kernel/random/poolsize is now 256 instead of 4096 bits,
> which was already small before.

The pool size for an RPNG is only the size of the state, nothing else.
It does not in any way describe how much you could get out.

> Why was such a change allowed into stable?

Because upstream considered it important enough for their stable
release, aka it fixes something important.

> This also breaks rngd’s --fill-watermark option when not set to
> percent values. Another reason this should not be changed within
> a stable series.

The kernel does not longer provide a number that could be used here.  It
might not have had any value before anyway.  You just need to reseed on
a regular interval.

Bastian

-- 
Ahead warp factor one, Mr. Sulu.



Bug#1013192: linux-image-5.10.0-15-amd64: ridiculously small entropy pool

2022-06-18 Thread Thorsten Glaser
Package: src:linux
Version: 5.10.120-1
Severity: serious
Tags: security
X-Debbugs-Cc: Debian Security Team 


/proc/sys/kernel/random/poolsize is now 256 instead of 4096 bits,
which was already small before.

Why was such a change allowed into stable?

This also breaks rngd’s --fill-watermark option when not set to
percent values. Another reason this should not be changed within
a stable series.


-- Package-specific info:
** Version:
Linux version 5.10.0-15-amd64 (debian-ker...@lists.debian.org) (gcc-10 (Debian 
10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2) #1 SMP 
Debian 5.10.120-1 (2022-06-09)

** Command line:
root=UUID=078df9a0-34f7-4171-b531-0cb628963204 ro clocksource=acpi_pm verbose

** Not tainted

** Kernel log:
Unable to read kernel log; any relevant messages should be attached

** Model information

** Loaded modules:
binfmt_misc
nfsd
auth_rpcgss
nfs_acl
nfs
lockd
grace
nfs_ssc
fscache
sunrpc
joydev
evdev
serio_raw
virtio_rng
rng_core
pcspkr
virtio_balloon
cirrus
drm_kms_helper
cec
drm
button
ext4
crc16
mbcache
jbd2
crc32c_generic
hid_generic
usbhid
hid
virtio_blk
virtio_net
net_failover
failover
ata_generic
crc32c_intel
psmouse
virtio_pci
virtio_ring
virtio
i2c_piix4
ata_piix
uhci_hcd
libata
floppy
ehci_hcd
scsi_mod
usbcore
usb_common

** PCI devices:
00:00.0 Host bridge [0600]: Intel Corporation 440FX - 82441FX PMC [Natoma] 
[8086:1237] (rev 02)
Subsystem: Red Hat, Inc. Qemu virtual machine [1af4:1100]
Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- 
Stepping- SERR+ FastB2B- DisINTx-
Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- TAbort- 
SERR- TAbort- 
SERR- TAbort- SERR- TAbort- 
SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- 
ii  firmware-linux-free  20200122-1

Versions of packages linux-image-5.10.0-15-amd64 suggests:
pn  debian-kernel-handbook   
pn  grub-pc | grub-efi-amd64 | extlinux  
pn  linux-doc-5.10   

Versions of packages linux-image-5.10.0-15-amd64 is related to:
pn  firmware-amd-graphics 
pn  firmware-atheros  
pn  firmware-bnx2 
pn  firmware-bnx2x
pn  firmware-brcm80211
pn  firmware-cavium   
pn  firmware-intel-sound  
pn  firmware-intelwimax   
pn  firmware-ipw2x00  
pn  firmware-ivtv 
pn  firmware-iwlwifi  
pn  firmware-libertas 
pn  firmware-linux-nonfree
pn  firmware-misc-nonfree 
pn  firmware-myricom  
pn  firmware-netxen   
pn  firmware-qlogic   
pn  firmware-realtek  
pn  firmware-samsung  
pn  firmware-siano
pn  firmware-ti-connectivity  
pn  xen-hypervisor

-- no debconf information