openbsd
19 months agoregister mvneta so the interface can be found by node/phandle later.
dlg [Mon, 3 Apr 2023 01:46:18 +0000 (01:46 +0000)]
register mvneta so the interface can be found by node/phandle later.

19 months agoadd phy_enable_prop_idx() to work with phys under props other than "phy".
dlg [Mon, 3 Apr 2023 01:40:32 +0000 (01:40 +0000)]
add phy_enable_prop_idx() to work with phys under props other than "phy".

eg, the snps,dwc3 device tree bindings say that it uses a usb2 and
usb3 phy, and they can be listed either "usb2-phy" and "usb3-phy"
under the standard "phys" and "phy-names" properties supported by
phy_enable(), or as slots 0 and 1 under a "usb-phy" properties. the
latter would be supported by phy_enable_idx(), but it hardcodes
"phys" as the property it looks at. phy_enable_prop_idx() is the
same as phy_enable_prop_idx, but it lets you specify which property
you're indexing into.

ok kettenis@

19 months agospecial case phandle 0 and return NULL when looking up network interfaces.
dlg [Mon, 3 Apr 2023 01:34:06 +0000 (01:34 +0000)]
special case phandle 0 and return NULL when looking up network interfaces.

not all interfaces will have a phandle (ie, it will be 0), so don't
let phandle 0 be used to find any of these.

discussed with miod@

19 months agoadd glue for network interfaces to be found by fdt/ofw node or phandle.
dlg [Mon, 3 Apr 2023 01:30:32 +0000 (01:30 +0000)]
add glue for network interfaces to be found by fdt/ofw node or phandle.

if we're going to support switch chips (eg, marvell link street
switches as found on a3700 boards like the espressobin), then the
device tree for switch ports identifies which network interface
they're connected by by a reference (phandle) across the device
tree. this lets network drivers register the ifnet struct with the
associated node and phandle so the switch can find it and configure
it for use with the switch.

ok kettenis@

19 months agodo actual init of the phy itself when needed.
dlg [Mon, 3 Apr 2023 01:21:31 +0000 (01:21 +0000)]
do actual init of the phy itself when needed.

turn the phy clock on and take the port out of suspend when something
(eg, xhci or ehci) wants to use the port.

before this the code just enabled power on the phy port regulators.

ok kettenis@

19 months agoclockintr: protect struct clockintr_queue with a mutex
cheloha [Mon, 3 Apr 2023 00:20:24 +0000 (00:20 +0000)]
clockintr: protect struct clockintr_queue with a mutex

Add a mutex (cq_mtx) to stuct clockintr_queue so that arbitrary CPUs
can manipulate clock interrupts established on arbitrary CPU queues.

Refactor the bulk of clockintr_schedule() into clockintr_schedule_locked()
so we can reuse it from within the mutex.

Tested by mlarkin@.  Neat bug found by mlarkin@.  With tweaks from
kettenis@.

ok kettenis@

19 months agofill out the list of quectel devices that should be supported.
dlg [Sun, 2 Apr 2023 23:57:57 +0000 (23:57 +0000)]
fill out the list of quectel devices that should be supported.

the list comes from the Quectel LTE&5G Linux USB Driver User Guide V2.0
and claims that all these devices function the same as each other.

ok miod@

19 months agoregen
dlg [Sun, 2 Apr 2023 23:56:19 +0000 (23:56 +0000)]
regen

19 months agoadd more quectel usb device ids
dlg [Sun, 2 Apr 2023 23:55:12 +0000 (23:55 +0000)]
add more quectel usb device ids

this list comes from the Quectel LTE&5G Linux USB Driver User Guide v2.0

ok miod@

19 months agoFix table by using strings of proper lengths instead of bogus
tb [Sun, 2 Apr 2023 23:33:42 +0000 (23:33 +0000)]
Fix table by using strings of proper lengths instead of bogus
scaling widths.

ok schwarze

19 months agoUse #define instead of `char * const' for the repeated GUID strings. Ugly,
miod [Sun, 2 Apr 2023 18:44:13 +0000 (18:44 +0000)]
Use #define instead of `char * const' for the repeated GUID strings. Ugly,
but restores compilability with older compilers such as gcc.

No intentional functional change.

19 months agoAdd support for TEMPerGold 3.4 to ugold(4).
miod [Sun, 2 Apr 2023 17:03:14 +0000 (17:03 +0000)]
Add support for TEMPerGold 3.4 to ugold(4).

ok landry@

19 months agoRegen
miod [Sun, 2 Apr 2023 17:02:08 +0000 (17:02 +0000)]
Regen

19 months agoNew TEMPerGold device.
miod [Sun, 2 Apr 2023 17:01:48 +0000 (17:01 +0000)]
New TEMPerGold device.

19 months agoadd Ryuichi Sakamoto. rest in peace.
sthen [Sun, 2 Apr 2023 16:43:42 +0000 (16:43 +0000)]
add Ryuichi Sakamoto. rest in peace.

19 months agoRevert r1.9 and reinstate r1.6
tb [Sun, 2 Apr 2023 15:36:53 +0000 (15:36 +0000)]
Revert r1.9 and reinstate r1.6

The argument change to x5519_ge_scalarmult_base() was made to match the
prototype in the header. More recent compilers warn about such ptr vs
array mismatches.

19 months agoadd Windows 11 _OSI strings
jsg [Sun, 2 Apr 2023 11:32:48 +0000 (11:32 +0000)]
add Windows 11 _OSI strings
ok mlarkin@

19 months agoAdd support for the RTL8211F-VD PHY.
kettenis [Sun, 2 Apr 2023 11:28:23 +0000 (11:28 +0000)]
Add support for the RTL8211F-VD PHY.

ok kevlo@

19 months agozap excessive semicolons; ok mglocker@
anton [Sun, 2 Apr 2023 08:53:01 +0000 (08:53 +0000)]
zap excessive semicolons; ok mglocker@

19 months agosync
kevlo [Sun, 2 Apr 2023 03:41:38 +0000 (03:41 +0000)]
sync

19 months agonable ngbe(4) on install media. Passes 'make release' build.
kevlo [Sun, 2 Apr 2023 03:40:54 +0000 (03:40 +0000)]
nable ngbe(4) on install media.  Passes 'make release' build.

Pointed out by miod@, patrick@

19 months agovmd(8): migrate vmd_vm.vm_ttyname to char array.
dv [Sun, 2 Apr 2023 02:04:10 +0000 (02:04 +0000)]
vmd(8): migrate vmd_vm.vm_ttyname to char array.

Other structs use a fixed length array already. This allows a vmd_vm
object to be transmitted over an ipc channel, too.

Additionally, solves a segfault caused by a strlcpy(3) in an error
path.

ok mlarkin@

19 months agomanpage for rkusbphy(4)
dlg [Sun, 2 Apr 2023 01:40:52 +0000 (01:40 +0000)]
manpage for rkusbphy(4)

19 months agoadd rkusbphy(4), a driver for the usb2phy on rockchip SoCs.
dlg [Sun, 2 Apr 2023 01:21:39 +0000 (01:21 +0000)]
add rkusbphy(4), a driver for the usb2phy on rockchip SoCs.

the rkusbphy device has children nodes in the device tree which are
the actual phys. this driver mostly exists to wire those children
up as PHYs and turn the associated regulators on when a host
controller enables the PHYs. in the future it should enable clocks
and take ports out of suspend too.

i'm not enabling this yet because it's useless without some tweaks
in the usb host controller drivers.

ok kettenis@

19 months agoRename get_softraid_{disks -> chunk}() and CRYPTO{DISK -> CHUNK}
kn [Sun, 2 Apr 2023 00:58:07 +0000 (00:58 +0000)]
Rename get_softraid_{disks -> chunk}() and CRYPTO{DISK -> CHUNK}

"disk" can mean both "chunk" and "volume" and a future diff for better
softraid volumes handling will benefit from this distinction.

No functional change.

19 months agotc_init.9: miscellaneous cleanup and rewrites
cheloha [Sun, 2 Apr 2023 00:02:26 +0000 (00:02 +0000)]
tc_init.9: miscellaneous cleanup and rewrites

- In DESCRIPTION, try to more fully describe what kern_tc.c does.
  Clean up the wording.

- Mention *all* the requirements for timekeeping hardware.  Describe
  the rollover margin in plainer language.

- Revise field descriptions for struct timecounter.  Don't mention
  fields the driver doesn't need to initialize.  Document the tc_user
  field.

- Add a CONTEXT section.

- In SEE ALSO, switch to an https URI on the main freebsd.org website.

- In HISTORY, note that the timecounting code first reached end users
  in FreeBSD 3.0.  This commit is probably the first one:

"Replace TOD clock code with more systematic approach."
https://cgit.freebsd.org/src/commit/sys/sys/timetc.h?id=7ec73f64179417aeda085c1c338385559fb49c23

- Add an AUTHORS section.

With input from Poul-Henning Kamp.

Link: https://marc.info/?l=openbsd-tech&m=168004968214914&w=2
ok jmc@

19 months agoLimit IFS scope to effect parsing code only
kn [Sat, 1 Apr 2023 23:24:46 +0000 (23:24 +0000)]
Limit IFS scope to effect parsing code only

Set to split `hw.disknames' strings is neat, but functionn-wide means the
bsort() invocation honours it, i.e. it'll output newlines not spaces.

Break the one-liner into the same multi-line idom occuring elsewhere and
contain IFS in the subshell that needs it.

This was never visible due how the output of was consumed by the shell.

The only visual, but otherwise effectless bug due to this appeared with the
new disk encryption question double-quoting get_dkdevs() output, i.e.
multiple disks printed across multiple lines;  this yields one, as inteded.

OK afresh1

19 months agoCheck all chunks in the delayed free list for write-after-free.
otto [Sat, 1 Apr 2023 18:47:51 +0000 (18:47 +0000)]
Check all chunks in the delayed free list for write-after-free.
Should catch more of them and closer (in time) to the WAF.  ok tb@

19 months agoUse char * const for the rest of the repeated GUID strings. More
krw [Sat, 1 Apr 2023 14:18:29 +0000 (14:18 +0000)]
Use char * const for the rest of the repeated GUID strings. More
readable, less prone to typos.

No intentional functional change.

19 months agoPull static const data out of BN_value_one()
tb [Sat, 1 Apr 2023 12:44:56 +0000 (12:44 +0000)]
Pull static const data out of BN_value_one()

Also use C99 initializers for readability.

discussed with jsing

19 months agoIndent labels
tb [Sat, 1 Apr 2023 11:10:55 +0000 (11:10 +0000)]
Indent labels

19 months agoGroup the non-constant time gcd functions together
tb [Sat, 1 Apr 2023 11:08:43 +0000 (11:08 +0000)]
Group the non-constant time gcd functions together

The only consumer of euclid() is BN_gcd(), which, in turn is only
used by BN_gcd_nonct(). Group them together rather than having
parts of the constant time implementation separate them.

This moves two functions to a different place in the file.

19 months agorkiovd(4)
kettenis [Sat, 1 Apr 2023 08:48:04 +0000 (08:48 +0000)]
rkiovd(4)

19 months agoEnable rkiovd(4)
kettenis [Sat, 1 Apr 2023 08:39:54 +0000 (08:39 +0000)]
Enable rkiovd(4)

19 months agoAdd rkiovd(4), a driver for the IO voltage domains on Rockchip SoCs. This
kettenis [Sat, 1 Apr 2023 08:39:05 +0000 (08:39 +0000)]
Add rkiovd(4), a driver for the IO voltage domains on Rockchip SoCs.  This
driver makes sure the SoC IO voltage domains stay in sync with the voltage
provided by the regulator associated with the domain.

ok dlg@

19 months agoImplement regulator notifiers which get called when the voltage/current
kettenis [Sat, 1 Apr 2023 08:37:23 +0000 (08:37 +0000)]
Implement regulator notifiers which get called when the voltage/current
for a regulator is changed or when the regulator gets initialized when it
attaches for the first time.  The latter makes it possible to register
a notifier for a regulator that hasn't attached yet.

ok dlg@

19 months agoavoid use of uninitialised memory
jsg [Sat, 1 Apr 2023 06:39:03 +0000 (06:39 +0000)]
avoid use of uninitialised memory
ok mlarkin@

19 months agohorrible whitespace, mostly on non-code lines. no object change
tb [Sat, 1 Apr 2023 05:27:44 +0000 (05:27 +0000)]
horrible whitespace, mostly on non-code lines. no object change

19 months agofollow quectel guidance on which usb interfaces umsm should match.
dlg [Sat, 1 Apr 2023 00:04:40 +0000 (00:04 +0000)]
follow quectel guidance on which usb interfaces umsm should match.

the Quectel LTE&5G Linux USB Driver User Guide V2.0 says umsm should
only attach to usb interfaces 0 to 3 using the interface class
UICLASS_VENDOR. their doco uses magic numbers, but this is what
they mean.

interfaces 4 and above provide network (not serial) via qmi, ecm,
or mbim. preventing umsm from attaching to the high interfaces
allows the appropriate network driver to use it instead. eg, umb
is now able to attach to the network interface because it presents
a standard mbim class.

discussed with and tested by kevlo@
ok patric@ sthen@ kevlo@

19 months agoshuffle the code in umsm_match a bit.
dlg [Fri, 31 Mar 2023 23:55:45 +0000 (23:55 +0000)]
shuffle the code in umsm_match a bit.

if umsm_lookup doesnt return anything, return early and leave the
rest of umsm_match to handling specific devices.

no functional change.

19 months agodont match quectel ec25 by vendor+product id
dlg [Fri, 31 Mar 2023 23:53:49 +0000 (23:53 +0000)]
dont match quectel ec25 by vendor+product id

quectel seem to provide a sane and consistent set of functionality
built on top of the qualcomm qmi stuff. their linux drivers guide
says quectel modems provide a set of umsm usb interfaces and a
network interface that can be in qmi, ecm, or mbim mode.

if the modem is in mbim mode, it will present the mbim classes which
umb should be able to attach to without this explicit vendor+product
match (assuming umsm doesn't claim the interface first).

based on info in the Quectel LTE&5G Linux USB Driver User Guide V2.0

discussed with and tested by kevlo@
ok patrick@ sthen@ kevlo@

19 months agoregen after "BE-M1000" addition
kn [Fri, 31 Mar 2023 20:31:35 +0000 (20:31 +0000)]
regen after "BE-M1000" addition

19 months agoAdd "Baikal Electronics" and their "BE-M1000" SoC
kn [Fri, 31 Mar 2023 20:28:48 +0000 (20:28 +0000)]
Add "Baikal Electronics" and their "BE-M1000" SoC

OK kettenis

19 months agoAdd missing NULL check after group_get()
tb [Fri, 31 Mar 2023 20:16:55 +0000 (20:16 +0000)]
Add missing NULL check after group_get()

Otherwise dh_getlen() will dereference ie->group and crash.

looks correct to hshoexer

19 months agoEnable ufshci(4).
mglocker [Fri, 31 Mar 2023 19:50:45 +0000 (19:50 +0000)]
Enable ufshci(4).

ok kettenis@

19 months agoFix white space.
bluhm [Fri, 31 Mar 2023 19:43:32 +0000 (19:43 +0000)]
Fix white space.

19 months agoAdd regress coverage for the new behavior of BN_copy() with respect to
tb [Fri, 31 Mar 2023 19:40:08 +0000 (19:40 +0000)]
Add regress coverage for the new behavior of BN_copy() with respect to
flags.

19 months agoCopy BN_FLG flags in BN_copy()
tb [Fri, 31 Mar 2023 19:39:15 +0000 (19:39 +0000)]
Copy BN_FLG flags in BN_copy()

BN_copy() forgot to copy the flags from the source to the target. Fix
this by copying the flags. In fact, only copy BN_FLG_CONSTTIME since
propagating BN_FLG_MALLOCED and BN_FLG_STATIC_DATA is wrong. Ignore the
BN_FLG_FREE flag "used for debugging" which of course means "unused"
like a lot of other debug code that somehow ended up in public headers.

Also: make BN_FLG_CONSTTIME sticky on the target, i.e., don't clear the
flag when copying from a non-constant time BIGNUM to a constant time one
for the following reason: if a is constant time, BN_sqr(a, a, ctx) would
use a BIGNUM without the flag internally, then copy the result to a in
which process a would lose its constant time flag.

Fixing this would be a lot of pointless work since someone had the good
sense of not relying on a fragile flag for something this important.
Rather, libcrypto always uses the constant time paths instead of the
faster, cryptographically inadequate paths.

Before this was changed, this was a pretty bad bug. The RSA code uses the
horrible BN_with_flags() function to create local versions of the private
moduli and set BN_FLG_CONSTTIME on them. If the RSA_FLAG_CACHE_PRIVATE for
caching moduli is set on the RSA, which it is by default, it attempts to
set these constant time versions on the RSA's internal Montgomery contexts.
Since it is called BN_MONT_CTX_set(), the setter doesn't set a BIGNUM on
the BN_MONT_CTX, rather it copies it over, losing the BN_FLG_CONSTTIME flag
in the process and make all the horrible leaky RSA code leak some more.
Good job.

This is all harmless and is mostly a cosmetic fix. BN_FLG_CONSTTIME should
be removed internally. It will be kept since various language bindings of
course picked it up and expose it.

ok beck jsing

19 months agoRefactor partition type menu code, disentangling mbr and gpt
krw [Fri, 31 Mar 2023 19:12:32 +0000 (19:12 +0000)]
Refactor partition type menu code, disentangling mbr and gpt
partition types from menu building and display.

Some GPT partition names change to match their MBR bretheren.

No intentional functional change.

19 months agoSince all arm64 assembly code in libc uses the ENTRY* macros from
kettenis [Fri, 31 Mar 2023 18:49:43 +0000 (18:49 +0000)]
Since all arm64 assembly code in libc uses the ENTRY* macros from
<machine/asm.h> they are already get the necessary "bti c" instructions.
Passi the -mmark-bti-property option to mark the corresponding object
files as having BTI support.

ok deraadt@

19 months agoAdd "bti c" where needed for BTI control flow integrety checks.
kettenis [Fri, 31 Mar 2023 18:46:24 +0000 (18:46 +0000)]
Add "bti c" where needed for BTI control flow integrety checks.

ok deraadt@

19 months agoCall entry point of the executable through register x17. This allows it
kettenis [Fri, 31 Mar 2023 18:45:05 +0000 (18:45 +0000)]
Call entry point of the executable through register x17.  This allows it
to be a normal C function that starts with "bti c" when BTI contro flow
integretry enforcement is in place.  Add "bti c" to _dl_bind_start().

Remove unused _rtld_tlsdesc() function to avoid having to add "bti c" to it.

ok deraadt@

19 months agoInline only use of TS_VERIFY_CTX_init()
tb [Fri, 31 Mar 2023 17:47:39 +0000 (17:47 +0000)]
Inline only use of TS_VERIFY_CTX_init()

Since TS_VERIFY_CTX is now opaque, the only thing TS_VERIFY_CTX_init()
is good for outside the library is memory leaks. Inside the library it's
also useless, since as a much more familiar name is memset(). It will soon
be able to join all the other nonsense that should never have leaked out of
this library.

19 months agoOops. Missed eliminating unneeded double quotes in previous.
krw [Fri, 31 Mar 2023 14:41:08 +0000 (14:41 +0000)]
Oops. Missed eliminating unneeded double quotes in previous.

19 months agoremove the whitespace in weird " / " constructs;
jmc [Fri, 31 Mar 2023 13:48:34 +0000 (13:48 +0000)]
remove the whitespace in weird " / " constructs;

19 months agoremove some unneccessary words and whitespace;
jmc [Fri, 31 Mar 2023 13:45:13 +0000 (13:45 +0000)]
remove some unneccessary words and whitespace;

19 months agoregen
kettenis [Fri, 31 Mar 2023 13:37:41 +0000 (13:37 +0000)]
regen

19 months agoAdd RTL8211F-VD, a new RTL8211F variant.
kettenis [Fri, 31 Mar 2023 13:37:02 +0000 (13:37 +0000)]
Add RTL8211F-VD, a new RTL8211F variant.

ok jsg@, dlg@

19 months agoBe more consistent in function naming. Functions taking or
krw [Fri, 31 Mar 2023 13:11:40 +0000 (13:11 +0000)]
Be more consistent in function naming. Functions taking or
returning struct uuid's use 'uuid' in their names.

Lengthen a pointlessly short line.

No functional change.

19 months agoremove unused unp_lock
jsg [Fri, 31 Mar 2023 12:35:24 +0000 (12:35 +0000)]
remove unused unp_lock
ok kn@ mvs@

19 months agoFlip label separators to fix previous
kn [Fri, 31 Mar 2023 12:07:54 +0000 (12:07 +0000)]
Flip label separators to fix previous

-gpiokeys0 at mainbus0, "PWR Button"
+gpiokeys0 at mainbus0: "PWR Button"

OK patrick

19 months agoWhen the slowcgi timeout fires but no process what yet started terminate
claudio [Fri, 31 Mar 2023 09:55:39 +0000 (09:55 +0000)]
When the slowcgi timeout fires but no process what yet started terminate
the session with a HTTP 408 error.
OK tb@

19 months agoEnable ngbe(4).
kevlo [Fri, 31 Mar 2023 08:22:09 +0000 (08:22 +0000)]
Enable ngbe(4).

ok miod@

19 months agoMan page for ngbe(4).
kevlo [Fri, 31 Mar 2023 08:20:58 +0000 (08:20 +0000)]
Man page for ngbe(4).

ok miod@

19 months agongbe(4), a driver for Wangxun WX1860 series Gigabit Ethernet devices.
kevlo [Fri, 31 Mar 2023 08:19:41 +0000 (08:19 +0000)]
ngbe(4), a driver for Wangxun WX1860 series Gigabit Ethernet devices.
Written based on the vendor driver for Linux.
Thanks to WangXun for relicensing their vendor driver to ISC.

Special thanks to miod@ for reviewing and providing valuable input.

ok miod@

19 months agoregen
kevlo [Fri, 31 Mar 2023 08:17:00 +0000 (08:17 +0000)]
regen

19 months agoAdd "Beijing WangXun Technology" vendor and WX1860 series NICs.
kevlo [Fri, 31 Mar 2023 08:16:30 +0000 (08:16 +0000)]
Add "Beijing WangXun Technology" vendor and WX1860 series NICs.

ok miod@

19 months agoGuard use of GROUP_EC2N with #ifndef OPENSSL_NO_EC2M
tb [Fri, 31 Mar 2023 07:28:46 +0000 (07:28 +0000)]
Guard use of GROUP_EC2N with #ifndef OPENSSL_NO_EC2M

This allows compiling isakmpd with a libcrypto that has binary field
support removed. Leave the enum value itself unguarded on claudio's
request.

ok beck claudio jsing

19 months agoAdd a regress for the recent BIO_new_NDEF() write after free
tb [Fri, 31 Mar 2023 06:07:44 +0000 (06:07 +0000)]
Add a regress for the recent BIO_new_NDEF() write after free

This is a simple reproducer for a write after free that avoids all the
mess with CMS, PKCS7 and SMIME. This now mostly allows ASAN to check
that the memory handling in this marvellous function is correct.

19 months agoExplicitly ignore return from waitpid here too.
dtucker [Fri, 31 Mar 2023 05:56:36 +0000 (05:56 +0000)]
Explicitly ignore return from waitpid here too.

19 months agoExplictly ignore return codes where we don't check them. From Dmitry
dtucker [Fri, 31 Mar 2023 04:45:08 +0000 (04:45 +0000)]
Explictly ignore return codes where we don't check them.  From Dmitry
Belyavskiy via github PR#238, ok djm@

19 months agoReturn immediately from get_sock_port if sock <0 so we don't call
dtucker [Fri, 31 Mar 2023 04:42:29 +0000 (04:42 +0000)]
Return immediately from get_sock_port if sock <0 so we don't call
getsockname on a negative FD.  From Coverity CID 291840, ok djm@

19 months agodon't leak arg2 on parse_pubkey_algos error path; ok dtucker@
djm [Fri, 31 Mar 2023 04:23:02 +0000 (04:23 +0000)]
don't leak arg2 on parse_pubkey_algos error path; ok dtucker@

19 months agoclamp max number of GSSAPI mechanisms to 2048; ok dtucker
djm [Fri, 31 Mar 2023 04:22:27 +0000 (04:22 +0000)]
clamp max number of GSSAPI mechanisms to 2048; ok dtucker

19 months agodon't print key if printing hostname failed; with/ok dtucker@
djm [Fri, 31 Mar 2023 04:21:56 +0000 (04:21 +0000)]
don't print key if printing hostname failed; with/ok dtucker@

19 months agoremove redundant test
djm [Fri, 31 Mar 2023 04:04:15 +0000 (04:04 +0000)]
remove redundant test

19 months agodon't attempt to decode a ridiculous number of attributes;
djm [Fri, 31 Mar 2023 04:00:37 +0000 (04:00 +0000)]
don't attempt to decode a ridiculous number of attributes;
harmless because of bounds elsewhere, but better to be explicit

19 months agoOnly send the used portion of struct idm_req in imsgs from the ldapclient
jmatthew [Fri, 31 Mar 2023 03:38:26 +0000 (03:38 +0000)]
Only send the used portion of struct idm_req in imsgs from the ldapclient
process to the main process. This significantly reduces memory usage when
updating larger directories.

ok aisha@

19 months agoremove unused variable; prompted by Coverity CID 291879
djm [Fri, 31 Mar 2023 03:22:49 +0000 (03:22 +0000)]
remove unused variable; prompted by Coverity CID 291879

19 months agoDocument i226.
kevlo [Fri, 31 Mar 2023 02:20:21 +0000 (02:20 +0000)]
Document i226.

From Brad.

19 months agodrm/amdkfd: Fix the memory overrun
jsg [Fri, 31 Mar 2023 02:16:10 +0000 (02:16 +0000)]
drm/amdkfd: Fix the memory overrun

From Ma Jun
8c31b663edc1cc5eb5c82282ab6ca99c69d1d942 in linux-6.1.y/6.1.22
4cc16d64b6cdb179a26fb389cae9dce788e88f5d in mainline linux

19 months agodrm/amdkfd: add GC 11.0.4 KFD support
jsg [Fri, 31 Mar 2023 02:14:03 +0000 (02:14 +0000)]
drm/amdkfd: add GC 11.0.4 KFD support

From Yifan Zhang
132f1d39439142547d9d5e45d7744d375dd21fdc in linux-6.1.y/6.1.22
88c21c2b56aa21dd34290d43ada74033dc3bfe35 in mainline linux

19 months agodrm/amdkfd: Fix the warning of array-index-out-of-bounds
jsg [Fri, 31 Mar 2023 02:12:21 +0000 (02:12 +0000)]
drm/amdkfd: Fix the warning of array-index-out-of-bounds

From Ma Jun
0ac954ec7f6c6dd6b7e94340599ba4b788d056d5 in linux-6.1.y/6.1.22
c0cc999f3c32e65a7c88fb323893ddf897b24488 in mainline linux

19 months agodrm/amdkfd: introduce dummy cache info for property asic
jsg [Fri, 31 Mar 2023 02:09:05 +0000 (02:09 +0000)]
drm/amdkfd: introduce dummy cache info for property asic

From Prike Liang
a5528973c676dad09a1af25cd49e4312238d1428 in linux-6.1.y/6.1.22
fd72e2cb2f9dd2734e8013b3e185a21f0d605d3e in mainline linux

19 months agodrm/amdgpu: reposition the gpu reset checking for reuse
jsg [Fri, 31 Mar 2023 02:07:08 +0000 (02:07 +0000)]
drm/amdgpu: reposition the gpu reset checking for reuse

From Tim Huang
75e10d7fb97906ddb641c7b09773a91693570f63 in linux-6.1.y/6.1.22
aaee0ce460b954e08b6e630d7e54b2abb672feb8 in mainline linux

19 months agodrm/amdgpu: skip ASIC reset for APUs when go to S4
jsg [Fri, 31 Mar 2023 02:04:27 +0000 (02:04 +0000)]
drm/amdgpu: skip ASIC reset for APUs when go to S4

From Tim Huang
9f7e1dd835ec834fb147a7725419497f0b0d83a7 in linux-6.1.y/6.1.22
b589626674de94d977e81c99bf7905872b991197 in mainline linux

19 months agodrm/i915: Preserve crtc_state->inherited during state clearing
jsg [Fri, 31 Mar 2023 02:02:47 +0000 (02:02 +0000)]
drm/i915: Preserve crtc_state->inherited during state clearing

From Ville Syrjala
220da3bf1fc335ae25b181d6bedcfc63efef6ccc in linux-6.1.y/6.1.22
3a84f2c6c9558c554a90ec26ad25df92fc5e05b7 in mainline linux

19 months agodrm/i915/active: Fix missing debug object activation
jsg [Fri, 31 Mar 2023 02:01:00 +0000 (02:01 +0000)]
drm/i915/active: Fix missing debug object activation

From Nirmoy Das
c355945957ef5e9bb05e0554fe4a6f92b0fcaf36 in linux-6.1.y/6.1.22
e92eb246feb9019b0b137706c934b8891cdfe3c2 in mainline linux

19 months agodrm/amdgpu/nv: Apply ASPM quirk on Intel ADL + AMD Navi
jsg [Fri, 31 Mar 2023 01:59:24 +0000 (01:59 +0000)]
drm/amdgpu/nv: Apply ASPM quirk on Intel ADL + AMD Navi

From Kai-Heng Feng
ed6364fb7c407ca53447d8b0723a61d320bcbae9 in linux-6.1.y/6.1.22
2b072442f4962231a8516485012bb2d2551ef2fe in mainline linux

19 months agodrm/amd/display: fix wrong index used in dccg32_set_dpstreamclk
jsg [Fri, 31 Mar 2023 01:56:10 +0000 (01:56 +0000)]
drm/amd/display: fix wrong index used in dccg32_set_dpstreamclk

From Hersen Wu
c590fce42c2117693bd0ee6646b3fc16a8f842ae in linux-6.1.y/6.1.22
4c94e57c258cb7800aa5f3a9d9597d91291407a9 in mainline linux

19 months agodrm/amd: Fix initialization mistake for NBIO 7.3.0
jsg [Fri, 31 Mar 2023 01:54:30 +0000 (01:54 +0000)]
drm/amd: Fix initialization mistake for NBIO 7.3.0

From Mario Limonciello
a0e39cdddc155926fc3c0395636b7c80e55e66a1 in linux-6.1.y/6.1.22
1717cc5f2962a4652c76ed3858b499ccae6c277c in mainline linux

19 months agodrm/amdgpu: Fix call trace warning and hang when removing amdgpu device
jsg [Fri, 31 Mar 2023 01:53:07 +0000 (01:53 +0000)]
drm/amdgpu: Fix call trace warning and hang when removing amdgpu device

From lyndonli
f06b902511ea05526f405ee64da54a8313d91831 in linux-6.1.y/6.1.22
93bb18d2a873d2fa9625c8ea927723660a868b95 in mainline linux

19 months agodrm/amd/display: Update clock table to include highest clock setting
jsg [Fri, 31 Mar 2023 01:51:13 +0000 (01:51 +0000)]
drm/amd/display: Update clock table to include highest clock setting

From Swapnil Patel
b66804fc7e21587431360af1dbcc1d9145237b7a in linux-6.1.y/6.1.22
2d99a7ec25cf456cd3680eb314d6454138e5aa64 in mainline linux

19 months agodrm/amd/display: Set dcn32 caps.seamless_odm
jsg [Fri, 31 Mar 2023 01:49:17 +0000 (01:49 +0000)]
drm/amd/display: Set dcn32 caps.seamless_odm

From Hersen Wu
612cf2495ec6cf5834e8b4a70f599190c35cf8c8 in linux-6.1.y/6.1.22
f9537b1fa7fb51c2162bc15ce469cbbf1ca0fbfe in mainline linux

19 months agodrm/i915/gt: perform uc late init after probe error injection
jsg [Fri, 31 Mar 2023 01:47:06 +0000 (01:47 +0000)]
drm/i915/gt: perform uc late init after probe error injection

From Andrzej Hajda
6e516595e88670baf275701893fd43e0a454f7d0 in linux-6.1.y/6.1.22
150784f9285e656373cf3953ef4a7663f1e1a0f2 in mainline linux

19 months agodrm/i915/guc: Fix missing ecodes
jsg [Fri, 31 Mar 2023 01:44:03 +0000 (01:44 +0000)]
drm/i915/guc: Fix missing ecodes

From John Harrison
be8c6ad6b16e6181510e35c53928381f0f39640a in linux-6.1.y/6.1.22
8df23e4c4f72f4e201c28e6fb0a67e2dbf30628a in mainline linux

19 months agodrm/i915/guc: Rename GuC register state capture node to be more obvious
jsg [Fri, 31 Mar 2023 01:42:14 +0000 (01:42 +0000)]
drm/i915/guc: Rename GuC register state capture node to be more obvious

From John Harrison
d2b9357d3da3f8054a47eef04c8ef361608b243e in linux-6.1.y/6.1.22
583ebae783b8241a30581c084ad6226051b594c5 in mainline linux

19 months agodrm/i915/fbdev: lock the fbdev obj before vma pin
jsg [Fri, 31 Mar 2023 01:39:50 +0000 (01:39 +0000)]
drm/i915/fbdev: lock the fbdev obj before vma pin

From Tejas Upadhyay
9595d71b2a274fe365240b53ce5d03c071c3e6e0 in linux-6.1.y/6.1.22
ed00eba03474adbf525ff03d69705d8c78b76456 in mainline linux

19 months agodrm/i915: Print return value on error
jsg [Fri, 31 Mar 2023 01:38:10 +0000 (01:38 +0000)]
drm/i915: Print return value on error

From Nirmoy Das
b6375c5ecd8c790e381ee3f9baeda8c9e5da4a45 in linux-6.1.y/6.1.22
20c68127e8e9d7899001c47465d0b79581f5fdc1 in mainline linux

19 months agodrm/amd/display: Fix DP MST sinks removal issue
jsg [Fri, 31 Mar 2023 01:36:11 +0000 (01:36 +0000)]
drm/amd/display: Fix DP MST sinks removal issue

From Cruise Hung
ee9caccc5e5cc1092baf6fb645525bba668b01c1 in linux-6.1.y/6.1.22
cbd6c1b17d3b42b7935526a86ad5f66838767d03 in mainline linux