CVS log for src/lib/libc/gen/arc4random.c

Up to [cvs.NetBSD.org] / src / lib / libc / gen

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Pull up following revision(s) (requested by riastradh in ticket #939):

	distrib/sets/lists/debug/module.md.amd64: revision 1.18
	sys/modules/Makefile: revision 1.292
	lib/libc/gen/arc4random.c: revision 1.34
	lib/libc/gen/arc4random.c: revision 1.35
	lib/libc/gen/arc4random.c: revision 1.36
	lib/libc/gen/arc4random.c: revision 1.37
	sys/kern/kern_entropy.c: revision 1.70
	lib/libc/gen/arc4random.c: revision 1.38
	sys/kern/kern_entropy.c: revision 1.71
	lib/libc/gen/getentropy.3: revision 1.8
	distrib/sets/lists/modules/md.amd64: revision 1.103
	share/man/man4/rnd.4: revision 1.42
	share/man/man4/rnd.4: revision 1.44
	lib/libc/include/arc4random.h: revision 1.1
	distrib/sets/lists/man/mi: revision 1.1786
	sys/arch/i386/conf/GENERIC: revision 1.1258
	sys/modules/acpivmgenid/acpivmgenid.ioconf: revision 1.1
	sys/arch/amd64/conf/ALL: revision 1.190
	distrib/sets/lists/debug/mi: revision 1.446
	sys/arch/i386/conf/ALL: revision 1.521
	lib/libc/gen/Makefile.inc: revision 1.219
	distrib/sets/lists/debug/module.md.i386: revision 1.12
	sys/dev/acpi/acpi_vmgenid.c: revision 1.1
	sys/dev/acpi/acpi_vmgenid.c: revision 1.2
	lib/libc/include/reentrant.h: revision 1.22
	sys/arch/evbarm/conf/GENERIC64: revision 1.219
	share/man/man4/Makefile: revision 1.735
	distrib/sets/lists/modules/md.i386: revision 1.100
	distrib/sets/lists/tests/mi: revision 1.1334
	lib/libc/gen/arc4random.3: revision 1.22
	sys/dev/acpi/files.acpi: revision 1.133
	lib/libc/gen/arc4random.3: revision 1.23
	tests/lib/libc/gen/t_arc4random.c: revision 1.1
	sys/sys/entropy.h: revision 1.6
	sys/arch/amd64/conf/GENERIC: revision 1.614
	sys/modules/acpivmgenid/Makefile: revision 1.1
	share/man/man4/acpivmgenid.4: revision 1.1
	lib/libc/gen/Makefile.inc: revision 1.220
	tests/lib/libc/gen/Makefile: revision 1.56
	share/man/man4/acpivmgenid.4: revision 1.2
	share/man/man4/acpivmgenid.4: revision 1.3

	(all via patch)

Catch up with all the lint warnings since exit on warning was disabled.

Disable 'missing header declaration' and 'nested extern' warnings for now.
acpivmgenid(4): New driver for virtual machine generation ID.

Added to amd64/ALL and i386/ALL kernel configurations, and made
available as a loadable module acpivmgenid.kmod on x86, for now.
TBD: Add to all ACPI-supporting GENERIC kernels.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

entropy(9): Factor out subroutines to reset and gather entropy.
`Reset' means we keep the data in the pool, but assume it had zero
entropy.  `Gather' means we request samples from all on-demand
sources and wait for the synchronous ones to complete.

No functional change intended, other than to expose new symbols --
just preparation to expose these to acpivmgenid(4), so it can use
these when the VM host notifies us that we, the guest, have been
cloned.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

acpivmgenid(4): Reset and gather entropy on VM clone notification.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random(3): Reseed if system entropy epoch changes.
This can happen, for example, if the system is a VM instance, and the
VM is cloned.

This incurs the cost of a system call on every arc4random call, which
is unfortunate, but
1. we don't currently have a (machine-independent) mechanism for
   exposing a read-only page to userland shared by the kernel to
   enable a cheaper access path to the entropy epoch; and
2. the algorithm here -- a simple application of ChaCha -- is likely
   also a bottleneck and could be much cheaper by
   (a) using sys/crypto/chacha for machine-dependent vectorized
       ChaCha code, and
   (b) filling a buffer (somewhere between a cipher block and a page)
       in a batch at a time, instead of running ChaCha to generate
       only 32 bytes at a time.
So although this might be a performance hit, the security benefit is
worthwhile and we have a clear path to do better than reversing the
performance hit later.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

acpivmgenid(4): Nix BUGS that have been squashed.
Reference kern.entropy.epoch for the remaining bug (which is a
performance issue, not a security issue).
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

entropy(9): Allow unprivileged reads of sysctl kern.entropy.epoch.

Applications need this in order to know when to reseed.  (We should
also expose it through a page shared read-only with userland for
cheaper access, but until we do, let's let applications get at it
through sysctl.)
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random.c: Fix test program.

This isn't wired up anywhere, but let's reduce the bitrot.  It was
helpful in reminding me that kern.entropy.epoch was, for reasons I
can't remember, restricted to privileged access.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

amd64, evbarm, i386: Add acpivmgenid(4) to GENERIC.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

rnd(4): Document kern.entropy.epoch is unprivileged and elaborate.
Cross-reference acpivmgenid(4).
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random(3): Note that arc4random respects kern.entropy.epoch.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork
Add debug info for new acpivmgenid module

arc4random(3): Add automatic tests.

This verifies that:
- arc4random zeroes its state and reseeds itself on fork
- arc4random reseeds itself on entropy consolidation (e.g., VM clone)
- arc4random falls back to global state if it can't allocate local
  state because address space limits cause mmap to fail

NOTE: This adds a new libc symbol __arc4random_global, but it's in
the reserved namespace and only used by t_arc4random, so no libc
minor bump.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

getentropy(3): Note intent to reseed on VM clone, and caveats.

Tidy markup and pacify some mandoc -Tlint complaints while here.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

Bump dates on man pages recently updated to mention VM clones.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random(3): Pacify some of lint's complaints.
PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random: suppress another lint warning

arc4random(3): Pacify some of lint's complaints.

PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random(3): Add automatic tests.

This verifies that:
- arc4random zeroes its state and reseeds itself on fork
- arc4random reseeds itself on entropy consolidation (e.g., VM clone)
- arc4random falls back to global state if it can't allocate local
  state because address space limits cause mmap to fail

NOTE: This adds a new libc symbol __arc4random_global, but it's in
the reserved namespace and only used by t_arc4random, so no libc
minor bump.

PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random.c: Fix test program.

This isn't wired up anywhere, but let's reduce the bitrot.  It was
helpful in reminding me that kern.entropy.epoch was, for reasons I
can't remember, restricted to privileged access.

PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

arc4random(3): Reseed if system entropy epoch changes.

This can happen, for example, if the system is a VM instance, and the
VM is cloned.

This incurs the cost of a system call on every arc4random call, which
is unfortunate, but

1. we don't currently have a (machine-independent) mechanism for
   exposing a read-only page to userland shared by the kernel to
   enable a cheaper access path to the entropy epoch; and

2. the algorithm here -- a simple application of ChaCha -- is likely
   also a bottleneck and could be much cheaper by

   (a) using sys/crypto/chacha for machine-dependent vectorized
       ChaCha code, and

   (b) filling a buffer (somewhere between a cipher block and a page)
       in a batch at a time, instead of running ChaCha to generate
       only 32 bytes at a time.

So although this might be a performance hit, the security benefit is
worthwhile and we have a clear path to do better than reversing the
performance hit later.

PR kern/58632: getentropy(2) and arc4random(3) do not reseed on VM
fork

Catch up with all the lint warnings since exit on warning was disabled.
Disable 'missing header declaration' and 'nested extern' warnings for now.

lib: remove CONSTCOND comment

Since 2021-01-31, lint doesn't need it anymore for the common pattern of
'do ... while (0)'.

Mostly merge changes from HEAD upto 20200411

lint is not smart enough to figure out that ilog2() is constant.

KNF

Missing MAP_PRIVATE on mmap

Remove #ifdef MAP_INHERIT_ZERO.

This is essential for fork-safety, so don't merely #warn about it.
Attaining fork-safety without it requires restructuring things -- in
particular, there's no clear way to make it per-thread and fork-safe
without some global list of states to zero on fork.

Pull up following revision(s) (requested by riastradh in ticket #597):
	lib/libc/gen/arc4random.c: revisions 1.26-1.28
	lib/libc/gen/arc4random.3:: revisions 1.10-1.19
Rewrite arc4random(3) with ChaCha20-based PRNG and per-thread state.
Explain the security model in the man page.
No more RC4!
--
Grammar.
--
Note relation of arc4random(3) to rand(3)/random(3).
--
Ruminate on security model choices and API design in arc4random(3).
--
Amplify comment about how quickly RC4 was known to be bad.
--
Markup for BUGS note about arc4random_uniform.
--
Cross-reference rnd(4).
--
Remove unnecessary Ns before punctuation. Fix a line.
--
Fix Google Groups link.
--
Tweak wording, define `output', remove misplaced scaremongering.
--
Fix non _REENTRANT build.
--
Use ChaCha20 here as advertised, not ChaCha8.
Oops.
Fortunately, there is no public cryptanalysis even of ChaCha8: the
best published attack is on ChaCha7 with time complexity 2^248.

Use ChaCha20 here as advertised, not ChaCha8.

Oops.

Fortunately, there is no public cryptanalysis even of ChaCha8: the
best published attack is on ChaCha7 with time complexity 2^248.

Fix non _REENTRANT build.

Rewrite arc4random(3) with ChaCha20-based PRNG and per-thread state.

Explain the security model in the man page.

No more RC4!

XXX pullup to netbsd-6, netbsd-5

Rebase to HEAD as of a few days ago.

Rebase.

Document that our use of pthread_mutex_unlock(3) is async-signal-safe
and as such can be used in a phtread_atfork(3) child handler.

fix missing backslash in previous

Wrap complex macros in do { ... } while (0).  Also replace the magic
number 1600000 with a macro.

Re-stir after forking, fixes PR lib/25367.
Re-stir after consuming 1600000 bytes, fixes PR lib/45952.

sync with head.

for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.

this commit was splitted into small chunks to avoid
a limitation of cvs.  ("Protocol error: too many arguments")

remove always inline because new gcc bitches.

sync with head

I'm fairly sure the libc rules require that arc4random_addrandom,
arc4random_buf, arc4random_stir and arc4random_uniform be weak.

arc4 is a random number sequence, there is no point using its own
output values to determine a number of output values to skip.
Skipping values on any possibly random event might be worth while, as
  might using the keying algorithm to stir in a possibly random value.

Since 'rs' is statically initialised (not bss) its s[] might as well be
  compile-time initialised as well.
arc4_init) is the same as arc4_stir().
Initialise rs.i to 0xff (not 0) so that the first key byte is processed
  with rs.i == 0 without the splurios rs.i-- in arc4_addrandom().
Remove the assignment rs.j = rs.i at the end of arc4_addrandom(), it isn't
  necessary and I can see no justificationm for it.
Replace RSIZE with __arraycount(as->s), however it is manifestly 256 (more
  correctly the number of values in rs.i and rs.j).

cache rs.i and rs.j in arc4random_buf() since the compiler can't be
  told that buf[] dosn't alias them.
Reduces the number of instructions inside the loop.
All the other functions are ok.

Change the logic used for thread locking to make the code more readable.
The effect on the object code is to replace a function call with a
  branch that will predict correctly in the non-threaded case.
Minor optimise of some other paths.

Use an inline function to check for initialisation, and an non-inlined
  one to do the actual initialise.
Fixes lib/46751 by removing all the replicated tests in each function.

Make this compile with the compiler I'm using.
Move variable defs to top of function.
Don't use const static mib[] - run time initialisation won't matter,
and not using static data may actually help in a .so.

sync with head

Pull up following revision(s) (requested by tls in ticket #92):
	lib/libc/gen/arc4random.c: revision 1.11
	lib/libc/gen/arc4random.c: revision 1.12
Make arc4random far less greedy for entropy.  Make arc4random actually
implement arc4 when used by threaded programs.
Fix bug in previous: don't reinitialize on every call!  Fix arc4random_buf so it actually ever initializes -- a security problem with revisions 1.9 and 1.10.

misc cleanups:
- const for mibs
- #define for magic constants
- casts

Fix bug in previous: don't reinitialize on every call!  Fix arc4random_buf so it actually ever initializes -- a security problem with revisions 1.9 and 1.10.

Make arc4random far less greedy for entropy.  Make arc4random actually
implement arc4 when used by threaded programs.

Sync with HEAD

add arc4random_{buf,uniform}, from OpenBSD.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Add missing __RCSID()

A little libc namespace housekeeping exercise:
* Make vfprintf_unlocked() an internal function, c.f. __svfscanf_unlocked().
* Add internal names for arc4random(), endnetpath(), fhstatvfs(),
  fstatvfs(), mkstemp(), shquote(), statvfs(), taddr2uaddr(), uaddr2taddr(),
  uuid_create_nil(), uuid_is_nil(), and wcwidth().
* Include namespace.h where supposed to.

Declare rs_initialized static.

Pullup rev 1.1-1.5 (requested by provos in ticket #1386)

Introduce bcrypt password scheme. Adds the arc4random API for creating
cryptographically strong random numbers.

Catch up to -current

file arc4random.c was added on branch netbsd-1-6 on 2004-03-26 22:52:50 +0000

Fix signed/unsigned comparison warnings.

Catch up to -current.

use sysctl(kern.urandom) if /dev/urandom is not present (like chroot jail)

Catch up to -current.

file arc4random.c was added on branch nathanw_sa on 2002-06-21 18:18:07 +0000

discard first 256 words when we stir.

/dev/arandom does not exist in netbsd.  use /dev/urandom.

add arc4random(3).  from openbsd