openbsd
16 months agodrm/amd/display: Fix Z8 support configurations
jsg [Tue, 13 Jun 2023 03:51:56 +0000 (03:51 +0000)]
drm/amd/display: Fix Z8 support configurations

From Nicholas Kazlauskas
97b3d8eed09785212900e9ccdc9ff1e0a86fef15 in linux-6.1.y/6.1.29
73dd4ca4b5a01235607231839bd351bbef75a1d2 in mainline linux

16 months agodrm/amd/display: Add debug option to skip PSR CRTC disable
jsg [Tue, 13 Jun 2023 03:50:21 +0000 (03:50 +0000)]
drm/amd/display: Add debug option to skip PSR CRTC disable

From Nicholas Kazlauskas
18225134088453e386b7aade107b3026d7aca76d in linux-6.1.y/6.1.29
00812bfc7bcb02faf127ee05f6ac27a5581eb701 in mainline linux

16 months agodrm/amd/display: Add Z8 allow states to z-state support list
jsg [Tue, 13 Jun 2023 03:48:04 +0000 (03:48 +0000)]
drm/amd/display: Add Z8 allow states to z-state support list

From Nicholas Kazlauskas
bcde2c8779932e21c42cf1797bef92651e9aa567 in linux-6.1.y/6.1.29
80676936805e46c79c38008e5142a77a1b2f2dc7 in mainline linux

16 months agodrm/amd/display: Refactor eDP PSR codes
jsg [Tue, 13 Jun 2023 03:46:07 +0000 (03:46 +0000)]
drm/amd/display: Refactor eDP PSR codes

From Ian Chen
83468820168a470a489e6ad6f7759d19687d1623 in linux-6.1.y/6.1.29
bd829d5707730072fecc3267016a675a4789905b in mainline linux

16 months agodrm/i915: Check pipe source size when using skl+ scalers
jsg [Tue, 13 Jun 2023 03:43:25 +0000 (03:43 +0000)]
drm/i915: Check pipe source size when using skl+ scalers

From Ville Syrjala
74a03d3c8d895a7d137bb4be8e40cae886f5d973 in linux-6.1.y/6.1.29
d944eafed618a8507270b324ad9d5405bb7f0b3e in mainline linux

16 months agodrm/i915/mtl: update scaler source and destination limits for MTL
jsg [Tue, 13 Jun 2023 03:41:25 +0000 (03:41 +0000)]
drm/i915/mtl: update scaler source and destination limits for MTL

From Animesh Manna
549ce5199d991eba81c477d5d05e988e2688abf7 in linux-6.1.y/6.1.29
f840834a8b60ffd305f03a53007605ba4dfbbc4b in mainline linux

16 months agodrm/amd: Use `amdgpu_ucode_*` helpers for MES
jsg [Tue, 13 Jun 2023 03:39:16 +0000 (03:39 +0000)]
drm/amd: Use `amdgpu_ucode_*` helpers for MES

From Mario Limonciello
7a0731130425562eac33c50697d3d25be283ef1f in linux-6.1.y/6.1.29
11e0b0067ec0707e8e598a5f9a547ab618ae7982 in mainline linux

16 months agodrm/amd: Add a new helper for loading/validating microcode
jsg [Tue, 13 Jun 2023 03:37:20 +0000 (03:37 +0000)]
drm/amd: Add a new helper for loading/validating microcode

From Mario Limonciello
a3e3a640d4fd9d7d40c1737e2b4373b7f4470eab in linux-6.1.y/6.1.29
2210af50ae7f4104269dfde7bafbbfbacdbe1a2b in mainline linux

16 months agodrm/amd: Load MES microcode during early_init
jsg [Tue, 13 Jun 2023 03:34:36 +0000 (03:34 +0000)]
drm/amd: Load MES microcode during early_init

From Mario Limonciello
3e1fa150e79897cc00db8348de267abf4c6c35be in linux-6.1.y/6.1.29
cc42e76e7de5190a7da5dac9d7b2bbb458e050bf in mainline linux

16 months agodrm/amdgpu: remove deprecated MES version vars
jsg [Tue, 13 Jun 2023 03:31:23 +0000 (03:31 +0000)]
drm/amdgpu: remove deprecated MES version vars

From Graham Sider
369b891842228e18821e17ce5dbafc99d37d8a5d in linux-6.1.y/6.1.29
6040517e4a29d3828160c571681eec9ffe10043f in mainline linux

16 months agodrm/amd/pm: avoid potential UBSAN issue on legacy asics
jsg [Tue, 13 Jun 2023 03:29:00 +0000 (03:29 +0000)]
drm/amd/pm: avoid potential UBSAN issue on legacy asics

From Guchun Chen
506da05a5e0fa46e048606581dd8bf3fe0161ab2 in linux-6.1.y/6.1.29
5247f05eadf1081a74b2233f291cee2efed25e3a in mainline linux

16 months agodrm/amdgpu: disable sdma ecc irq only when sdma RAS is enabled in suspend
jsg [Tue, 13 Jun 2023 03:27:33 +0000 (03:27 +0000)]
drm/amdgpu: disable sdma ecc irq only when sdma RAS is enabled in suspend

From Guchun Chen
2a179117a3b29e7136e4045c57090a05bb97f373 in linux-6.1.y/6.1.29
8b229ada2669b74fdae06c83fbfda5a5a99fc253 in mainline linux

16 months agodrm/amd/pm: parse pp_handle under appropriate conditions
jsg [Tue, 13 Jun 2023 03:26:00 +0000 (03:26 +0000)]
drm/amd/pm: parse pp_handle under appropriate conditions

From Guchun Chen
17a69415679c6e0b55b390e86d59de82d7e53b9e in linux-6.1.y/6.1.29
58d9b9a14b47c2a3da6effcbb01607ad7edc0275 in mainline linux

16 months agodrm/amd/display: Enforce 60us prefetch for 200Mhz DCFCLK modes
jsg [Tue, 13 Jun 2023 03:23:52 +0000 (03:23 +0000)]
drm/amd/display: Enforce 60us prefetch for 200Mhz DCFCLK modes

From Alvin Lee
348dcdf102a44ab5b389c4cd932bc1a61e210f6d in linux-6.1.y/6.1.29
b504f99ccaa64da364443431e388ecf30b604e38 in mainline linux

16 months agodrm/amdgpu: Fix vram recover doesn't work after whole GPU reset (v2)
jsg [Tue, 13 Jun 2023 03:21:05 +0000 (03:21 +0000)]
drm/amdgpu: Fix vram recover doesn't work after whole GPU reset (v2)

From Lin.Cao
7a8248317b322d3cb56c64332062bae91460978a in linux-6.1.y/6.1.29
6c032c37ac3ef3b7df30937c785ecc4da428edc0 in mainline linux

16 months agodrm/amdgpu: change gfx 11.0.4 external_id range
jsg [Tue, 13 Jun 2023 03:18:50 +0000 (03:18 +0000)]
drm/amdgpu: change gfx 11.0.4 external_id range

From Yifan Zhang
6197fb331a6e274355bbcd04386a2631e8cf7f1c in linux-6.1.y/6.1.29
996e93a3fe74dcf9d467ae3020aea42cc3ff65e3 in mainline linux

16 months agodrm/amdgpu/jpeg: Remove harvest checking for JPEG3
jsg [Tue, 13 Jun 2023 03:17:09 +0000 (03:17 +0000)]
drm/amdgpu/jpeg: Remove harvest checking for JPEG3

From Saleemkhan Jamadar
28c2e072fa1382e7e4da51ff7795fb5b5774f5a7 in linux-6.1.y/6.1.29
5b94db73e45e2e6c2840f39c022fd71dfa47fc58 in mainline linux

16 months agodrm/amdgpu/gfx: disable gfx9 cp_ecc_error_irq only when enabling legacy gfx ras
jsg [Tue, 13 Jun 2023 03:15:33 +0000 (03:15 +0000)]
drm/amdgpu/gfx: disable gfx9 cp_ecc_error_irq only when enabling legacy gfx ras

From Guchun Chen
f661ad53658a1ea35c004af1f5fbe25c4d1cdb08 in linux-6.1.y/6.1.29
4a76680311330aefe5074bed8f06afa354b85c48 in mainline linux

16 months agodrm/amdgpu: fix amdgpu_irq_put call trace in gmc_v11_0_hw_fini
jsg [Tue, 13 Jun 2023 03:13:20 +0000 (03:13 +0000)]
drm/amdgpu: fix amdgpu_irq_put call trace in gmc_v11_0_hw_fini

From Horatio Zhang
02e6cb9b3aeffc6b0e3955f6e0346293e2415cbc in linux-6.1.y/6.1.29
13af556104fa93b1945c70bbf8a0a62cd2c92879 in mainline linux

16 months agodrm/amdgpu: fix an amdgpu_irq_put() issue in gmc_v9_0_hw_fini()
jsg [Tue, 13 Jun 2023 03:11:26 +0000 (03:11 +0000)]
drm/amdgpu: fix an amdgpu_irq_put() issue in gmc_v9_0_hw_fini()

From Hamza Mahfooz
59cb2d46e177894c3554a25a3358b72d4bee31d3 in linux-6.1.y/6.1.29
922a76ba31adf84e72bc947267385be420c689ee in mainline linux

16 months agodrm/amdgpu: fix amdgpu_irq_put call trace in gmc_v10_0_hw_fini
jsg [Tue, 13 Jun 2023 03:09:53 +0000 (03:09 +0000)]
drm/amdgpu: fix amdgpu_irq_put call trace in gmc_v10_0_hw_fini

From Horatio Zhang
59e2439111ac2bd24ea0cecf5825cf06684b2c6c in linux-6.1.y/6.1.29
08c677cb0b436a96a836792bb35a8ec5de4999c2 in mainline linux

16 months agodrm/amd/display: fix flickering caused by S/G mode
jsg [Tue, 13 Jun 2023 03:07:52 +0000 (03:07 +0000)]
drm/amd/display: fix flickering caused by S/G mode

From Hamza Mahfooz
f2e43c98042c607376479484ea8f98d3452868f5 in linux-6.1.y/6.1.29
08da182175db4c7f80850354849d95f2670e8cd9 in mainline linux

16 months agodrm/amd/display: filter out invalid bits in pipe_fuses
jsg [Tue, 13 Jun 2023 03:05:48 +0000 (03:05 +0000)]
drm/amd/display: filter out invalid bits in pipe_fuses

From Samson Tam
4c1e747ca61c6103bcc259ad7f2fc6d53acff291 in linux-6.1.y/6.1.29
682439fffad9fa9a38d37dd1b1318e9374232213 in mainline linux

16 months agodrm/amd/display: Fix 4to1 MPC black screen with DPP RCO
jsg [Tue, 13 Jun 2023 03:04:00 +0000 (03:04 +0000)]
drm/amd/display: Fix 4to1 MPC black screen with DPP RCO

From Nicholas Kazlauskas
c2b2641ecb9aed1613976a2abf56292e206e3694 in linux-6.1.y/6.1.29
bf224e00a9f54e2bf14b4d720a09c3d2f4aa4aa8 in mainline linux

16 months agodrm/amd/display: Add NULL plane_state check for cursor disable logic
jsg [Tue, 13 Jun 2023 03:01:44 +0000 (03:01 +0000)]
drm/amd/display: Add NULL plane_state check for cursor disable logic

From Nicholas Kazlauskas
cc9942840afaa3cbbb90dd41404d15125ae5d192 in linux-6.1.y/6.1.29
d29fb7baab09b6a1dc484c9c67933253883e770a in mainline linux

16 months agodrm/i915/dsi: Use unconditional msleep() instead of intel_dsi_msleep()
jsg [Tue, 13 Jun 2023 02:59:36 +0000 (02:59 +0000)]
drm/i915/dsi: Use unconditional msleep() instead of intel_dsi_msleep()

From Hans de Goede
aa0f98c5d1962b4dedec00067fc1b28a6d4f7d65 in linux-6.1.y/6.1.29
c8c2969bfcba5fcba3a5b078315c1b586d927d9f in mainline linux

16 months agodrm/i915/color: Fix typo for Plane CSC indexes
jsg [Tue, 13 Jun 2023 02:56:18 +0000 (02:56 +0000)]
drm/i915/color: Fix typo for Plane CSC indexes

From Chaitanya Kumar Borah
5b6b81decdf08b5171c3872e45e9e3772a274032 in linux-6.1.y/6.1.29
2efc8e1001acfdc143cf2d25a08a4974c322e2a8 in mainline linux

16 months agodrm/amdgpu: add a missing lock for AMDGPU_SCHED
jsg [Tue, 13 Jun 2023 02:54:09 +0000 (02:54 +0000)]
drm/amdgpu: add a missing lock for AMDGPU_SCHED

From Chia-I Wu
7887397338a55a2439ba8db6d9a93cbe4094d912 in linux-6.1.y/6.1.29
2397e3d8d2e120355201a8310b61929f5a8bd2c0 in mainline linux

16 months agodrm/i915/mtl: Add the missing CPU transcoder mask in intel_device_info
jsg [Tue, 13 Jun 2023 02:52:19 +0000 (02:52 +0000)]
drm/i915/mtl: Add the missing CPU transcoder mask in intel_device_info

From Radhakrishna Sripada
4b08cdd239e7d15f11d2be07e59c3af478f61b11 in linux-6.1.y/6.1.29
6ece90e3665a9b7fb2637fcca26cebd42991580b in mainline linux

16 months agodrm/amd/display: Update bounding box values for DCN321
jsg [Tue, 13 Jun 2023 02:50:05 +0000 (02:50 +0000)]
drm/amd/display: Update bounding box values for DCN321

From Aurabindo Pillai
80a791a19902edc7c0fc7f8d2f3d411531e6a4ca in linux-6.1.y/6.1.29
989cd3e76a4aab76fe7dd50090ac3fa501c537f6 in mainline linux

16 months agodrm/amd/display: Do not clear GPINT register when releasing DMUB from reset
jsg [Tue, 13 Jun 2023 02:48:26 +0000 (02:48 +0000)]
drm/amd/display: Do not clear GPINT register when releasing DMUB from reset

From Aurabindo Pillai
7bba2e5e096e005f9ea5dba85a224caf9e80909f in linux-6.1.y/6.1.29
99d92eaca5d915763b240aae24669f5bf3227ecf in mainline linux

16 months agodrm/amd/display: Reset OUTBOX0 r/w pointer on DMUB reset
jsg [Tue, 13 Jun 2023 02:46:50 +0000 (02:46 +0000)]
drm/amd/display: Reset OUTBOX0 r/w pointer on DMUB reset

From Cruise Hung
ccb0ad946adc43d9b146323228a365aa7400fd42 in linux-6.1.y/6.1.29
425afa0ac99a05b39e6cd00704fa0e3e925cee2b in mainline linux

16 months agodrm/amd/display: Fixes for dcn32_clk_mgr implementation
jsg [Tue, 13 Jun 2023 02:45:05 +0000 (02:45 +0000)]
drm/amd/display: Fixes for dcn32_clk_mgr implementation

From Aurabindo Pillai
bb13726625e7d6220744fac823baec4ce9e7f563 in linux-6.1.y/6.1.29
d1c5c3e252b8a911a524e6ee33b82aca81397745 in mainline linux

16 months agoregen
jcs [Tue, 13 Jun 2023 02:44:08 +0000 (02:44 +0000)]
regen

16 months agoadd SanDisk PC SN740
jcs [Tue, 13 Jun 2023 02:43:39 +0000 (02:43 +0000)]
add SanDisk PC SN740

16 months agodrm/amd/display: Return error code on DSC atomic check failure
jsg [Tue, 13 Jun 2023 02:42:49 +0000 (02:42 +0000)]
drm/amd/display: Return error code on DSC atomic check failure

From Hersen Wu
b7ae53dd0d290b27fd4e0bdeff5849ecb3d588a3 in linux-6.1.y/6.1.29
dd24662d9dfbad281bbf030f06d68c7938fa0c66 in mainline linux

16 months agodrm/amd/display: Add missing WA and MCLK validation
jsg [Tue, 13 Jun 2023 02:40:37 +0000 (02:40 +0000)]
drm/amd/display: Add missing WA and MCLK validation

From Rodrigo Siqueira
374f7fa01ae56bc000dc1d54e80a8f4e7f606028 in linux-6.1.y/6.1.29
822b84ecfc646da0f87fd947fa00dc3be5e45ecc in mainline linux

16 months agodrm/amd/display: Remove FPU guards from the DML folder
jsg [Tue, 13 Jun 2023 02:38:27 +0000 (02:38 +0000)]
drm/amd/display: Remove FPU guards from the DML folder

From Rodrigo Siqueira
0b47019f544fbf1667798085b71374fe4d09e611 in linux-6.1.y/6.1.29
bbfbf09d193ac831c40db50ef4b31d11548a9eef in mainline linux

16 months agodrm/amd/display: Ext displays with dock can't recognized after resume
jsg [Tue, 13 Jun 2023 02:36:09 +0000 (02:36 +0000)]
drm/amd/display: Ext displays with dock can't recognized after resume

From Ryan Lin
28d0f85aff342921d75597bc4997f9d6e83be2ef in linux-6.1.y/6.1.29
1e5d4d8eb8c0f15d90c50e7abd686c980e54e42e in mainline linux

16 months agoMove nd6_ifdetach() out of netlock. In this point, the interface is
mvs [Mon, 12 Jun 2023 21:19:54 +0000 (21:19 +0000)]
Move nd6_ifdetach() out of netlock. In this point, the interface is
disconnected from everywhere. No need to hold netlock for dummy
'nd_ifinfo' release. Netlock is also not needed for
TAILQ_EMPTY(&ifp->if_*hooks) assertions.

ok kn bluhm

16 months agoSimple seq(1) regress. More tests are needed.
millert [Mon, 12 Jun 2023 20:19:45 +0000 (20:19 +0000)]
Simple seq(1) regress.  More tests are needed.

16 months agoseq: fix check for rounding error/truncation
millert [Mon, 12 Jun 2023 20:15:06 +0000 (20:15 +0000)]
seq: fix check for rounding error/truncation
We need to compare the printable version of the last value displayed,
not the floating point representation.  Otherwise, we may print the
last value twice.  OK deraadt@

16 months agoDowngrade CMS signing-time being after notAfter to a warning
job [Mon, 12 Jun 2023 18:22:02 +0000 (18:22 +0000)]
Downgrade CMS signing-time being after notAfter to a warning

Feedback from Ties, Ben

OK tb@

16 months agoRemove prototypes for various ec_GF2m_* functions that no longer exist.
jsing [Mon, 12 Jun 2023 18:17:18 +0000 (18:17 +0000)]
Remove prototypes for various ec_GF2m_* functions that no longer exist.

16 months agoOptimise quad word primitives on aarch64.
jsing [Mon, 12 Jun 2023 16:42:11 +0000 (16:42 +0000)]
Optimise quad word primitives on aarch64.

This provides a performance gain across most BN operations.

16 months agoProvide and use various quad word primitives.
jsing [Mon, 12 Jun 2023 16:17:24 +0000 (16:17 +0000)]
Provide and use various quad word primitives.

This includes bn_qwaddqw(), bn_qwsubqw(), bn_qwmulw_addw() and
bn_qwmulw_addqw_addw(). These can typically be optimised on architectures
that have a reasonable number of general purpose registers.

ok tb@

16 months agoReduce issues with types by switching iosz and totalsz from off_t to size_t.
claudio [Mon, 12 Jun 2023 15:27:52 +0000 (15:27 +0000)]
Reduce issues with types by switching iosz and totalsz from off_t to size_t.
In rpki-client the maximum file size is limited to 2GB so even on 32bit
archs size_t is large enough.
This solves some of the signed vs unsigned issues between bufsz/bufpos
and iosz.
OK tb@

16 months agoAdd content-encoding compression support (just gzip and deflate).
claudio [Mon, 12 Jun 2023 14:56:38 +0000 (14:56 +0000)]
Add content-encoding compression support (just gzip and deflate).
This will allow servers to send compressed XML which saves around 50%.
The uncompressed output is limited to MAX_CONTENTLEN bytes so the
impact of decompression bombs is limited.
With and OK job@ tb@

16 months agoUse attr_writebuf() instead of hand rolling a more complicated version
claudio [Mon, 12 Jun 2023 12:48:07 +0000 (12:48 +0000)]
Use attr_writebuf() instead of hand rolling a more complicated version
for IMSG_CTL_SHOW_RIB_ATTR. Also drop the attr_optlen() usage in
imsg_create() since it is not stricly needed. With this attr_optlen
follows the path of the dodo.
OK tb@

16 months agoUse data != NULL to be more explicit. No functional change.
claudio [Mon, 12 Jun 2023 12:10:17 +0000 (12:10 +0000)]
Use data != NULL to be more explicit. No functional change.
OK tb@

16 months agoadd Mercusys MW150US V2
jsg [Mon, 12 Jun 2023 11:30:55 +0000 (11:30 +0000)]
add Mercusys MW150US V2
tested by Daeil Lee

16 months agomatch Mercusys MW150US V2
jsg [Mon, 12 Jun 2023 11:27:30 +0000 (11:27 +0000)]
match Mercusys MW150US V2
from Daeil Lee

16 months agoregen
jsg [Mon, 12 Jun 2023 11:26:54 +0000 (11:26 +0000)]
regen

16 months agoadd Mercusys MW150US V2
jsg [Mon, 12 Jun 2023 11:26:24 +0000 (11:26 +0000)]
add Mercusys MW150US V2
from Daeil Lee

16 months agoUse stdio open_memstream(3) to build up log strings instead of trying to
claudio [Mon, 12 Jun 2023 09:02:31 +0000 (09:02 +0000)]
Use stdio open_memstream(3) to build up log strings instead of trying to
abuse ibufs for that. Using stdio for this has the benefit of using any
stdio function to build up strings including fprintf().
With and OK tb@

16 months agoregen
jsg [Mon, 12 Jun 2023 01:13:13 +0000 (01:13 +0000)]
regen

16 months agoadd more Navi 31 device ids
jsg [Mon, 12 Jun 2023 01:12:33 +0000 (01:12 +0000)]
add more Navi 31 device ids

Radeon Pro W7800 and Radeon Pro W7900
from Radeon Software for Linux 23.10.1 (5.5.1) libdrm-amdgpu-common

16 months agoDisable PAC with the architected algorithm for now, but leave it enabled
kettenis [Sun, 11 Jun 2023 21:42:01 +0000 (21:42 +0000)]
Disable PAC with the architected algorithm for now, but leave it enabled
when the hardware uses an implementation defined algorithm.  There are
issues with PAC on the x13s (but not on the windows dev kit) which uses
the architected algorithm as it uses a core designed by ARM.  This leaves
PAC enabled on Apple hardware.

ok deraadt@

16 months agoConvert legacy server kex to one-shot sign/verify
tb [Sun, 11 Jun 2023 19:01:01 +0000 (19:01 +0000)]
Convert legacy server kex to one-shot sign/verify

This converts ssl3_{get,send}_server_key_exchange() to EVP_DigestVerify()
and EVP_DigestSign(). In order to do this, build the full signed_params
up front and rework the way the key exchange parameters are constructed.
This way we can do the verify and sign steps in one go and at the same
use a more idiomatic approach with CBB/CBS.

with/ok jsing

16 months agoEasy EVP_Digest{Sign,Verify} conversions for legacy stack
tb [Sun, 11 Jun 2023 18:50:51 +0000 (18:50 +0000)]
Easy EVP_Digest{Sign,Verify} conversions for legacy stack

Convert ssl3_send_client_verify_{sigalgs,gost}() to EVP_DigestSign() and
ssl3_get_cert_verify() to EVP_DigestVerify().

ok jsing

16 months agoNo need to check for DTYPE_FLOPPY. If there is neither a GPT nor
krw [Sun, 11 Jun 2023 14:00:04 +0000 (14:00 +0000)]
No need to check for DTYPE_FLOPPY. If there is neither a GPT nor
an MBR then install biosboot in sector 0.

Without the check for DTYPE_FLOPPY there is no need for
FSDISKTYPE=floppy3 and therefore flip the last two Makefiles to
the "echo '/ *' | disklabel -wAT-" idiom.

Feedback/fix from kn@

16 months agoremove unused args_st struct
jsg [Sun, 11 Jun 2023 13:02:10 +0000 (13:02 +0000)]
remove unused args_st struct
ok tb@

16 months agoremove chopup_args() unused since apps.c rev 1.31
jsg [Sun, 11 Jun 2023 12:35:00 +0000 (12:35 +0000)]
remove chopup_args() unused since apps.c rev 1.31
ok tb@

16 months agoopenssl enc: drop a few parens and unwrap a few lines
tb [Sun, 11 Jun 2023 12:06:08 +0000 (12:06 +0000)]
openssl enc: drop a few parens and unwrap a few lines

No binary change on amd64

16 months agoopenssl enc: small style fixup after ZLIB unifdef
tb [Sun, 11 Jun 2023 11:54:44 +0000 (11:54 +0000)]
openssl enc: small style fixup after ZLIB unifdef

16 months agofix typo: 'hash buffer to small' -> too small
op [Sun, 11 Jun 2023 10:30:10 +0000 (10:30 +0000)]
fix typo: 'hash buffer to small' -> too small

16 months agoUnifdef ZLIB
tb [Sun, 11 Jun 2023 05:45:20 +0000 (05:45 +0000)]
Unifdef ZLIB

This is very dead code: the openssl app was never compiled with -DZLIB
after January 1, 2015.

16 months agoUnifdef ZLIB
tb [Sun, 11 Jun 2023 05:35:43 +0000 (05:35 +0000)]
Unifdef ZLIB

This has long been unused code and compilation with -DZLIB was broken
for a long time after BIO was made opaque.

ok jsing

16 months agoImplement support for pointer authentication (PAC) in userland. With PAC
kettenis [Sat, 10 Jun 2023 19:30:48 +0000 (19:30 +0000)]
Implement support for pointer authentication (PAC) in userland.  With PAC
it is possible to "sign" pointers with a hidden key.  The signature is
placed in unused bits of the pointer and can be checked later.  This can
be used to provide "tail CFI" that is similar to what retguard provides.

Debuggers need to be aware of the fact that pointers can be signed.  For
this purpose a new PT_PACMASK ptrace(2) request is introduced that returns
as mask that indicates the bits used for the signature.  Separate masks
are provided for code and data pointers even though the masks are identical
in the current implementation.  These masks are also written into a special
note section in the core dump.

ok patrick@

16 months agoAdd qcpas(4), a driver for the Peripheral Authentication Service found on
patrick [Sat, 10 Jun 2023 18:31:38 +0000 (18:31 +0000)]
Add qcpas(4), a driver for the Peripheral Authentication Service found on
Qualcomm SoCs.

The immediate task for this driver is to provide firmware to the auxiliary
cores and to bring them up.  This is accomplished by parsing the ELF files
and providing the data in certain memory regions, and telling qcscm(4) to
check and execute the firmware on the auxiliary cores.

With the cores up we can now talk to the firmware.  The glink-edge subnode
indicates that we can talk to it using the GLINK protocol over shared memory
provided by qcsmem(4).  This interface is essentially a channel multiplexer,
with each channel identified through an ASCII string.

One of those channels connects to a PMIC router, which allows us to talk to
the battery manager service that contains information about the charging and
battery states.

ok drahn@ kettenis@

16 months agoConvert EVP_Digest{Sign,Verify}* to one-shot for TLSv1.3
tb [Sat, 10 Jun 2023 15:34:36 +0000 (15:34 +0000)]
Convert EVP_Digest{Sign,Verify}* to one-shot for TLSv1.3

Using one-shot EVP_DigestSign() and EVP_DigestVerify() is slightly shorter
and is needed for Ed25519 support.

ok jsing

16 months agosync
deraadt [Sat, 10 Jun 2023 15:16:43 +0000 (15:16 +0000)]
sync

16 months agoksh: remove broken special handling of test -t
op [Sat, 10 Jun 2023 07:24:21 +0000 (07:24 +0000)]
ksh: remove broken special handling of test -t

Drop the vestiges of the pre-POSIX support of `test -t' defaulting to fd
1.  It doesn't work and it always succeed since "-t" is treated as a
string by default when no argument (fd) is specified.

diff by Lucas (lucas [at] sexy [dot] is) with minor change by me.
ok millert@

16 months agotest: fix description of -t: it has no default
op [Sat, 10 Jun 2023 07:19:39 +0000 (07:19 +0000)]
test: fix description of -t: it has no default

-t always requires the fd number as argument, there's no default.  With
only one argument -t is equivalent to `test -n -t' and so banally always
true.

diff from Lucas (lucas [at] sexy [dot] is)
ok millert@

16 months agoFile new test-bleichenbacher-timing-pregenerate.py under failing tests
tb [Sat, 10 Jun 2023 05:00:58 +0000 (05:00 +0000)]
File new test-bleichenbacher-timing-pregenerate.py under failing tests
until someone finds time and motivation to figure out how to use this.

16 months agowe always create keys 2 releases into the future
deraadt [Fri, 9 Jun 2023 15:31:44 +0000 (15:31 +0000)]
we always create keys 2 releases into the future

16 months agoReadd "-wgpsk", accidentially dropped in r1.465 adding "wgdescr"
kn [Fri, 9 Jun 2023 12:22:01 +0000 (12:22 +0000)]
Readd "-wgpsk", accidentially dropped in r1.465 adding "wgdescr"

Noticed by Bradley Latus
Diff from Jane Johansson
OK tb

16 months agoRemove dead code.
beck [Thu, 8 Jun 2023 22:02:40 +0000 (22:02 +0000)]
Remove dead code.

must_be_ca can no longer be 0 after the proxy cert code got nuked,
so change this to an if. must_be_ca is now -1 for a leaf, or 1 for
a non leaf.

ok tb@

16 months agoFix mismatch between function prototype and definition, from Anindya
nicm [Thu, 8 Jun 2023 11:17:28 +0000 (11:17 +0000)]
Fix mismatch between function prototype and definition, from Anindya
Mukherjee.

16 months agoFrom the description of "openssl verify", delete the duplicate and
schwarze [Thu, 8 Jun 2023 09:40:17 +0000 (09:40 +0000)]
From the description of "openssl verify", delete the duplicate and
outdated list of error messages.  Instead, refer to the master copy
of that list in X509_STORE_CTX_get_error(3).

Suggested by and OK tb@, and beck@ also agrees with the idea.

16 months agoadd tests related to --libs-only-l and the likes
espie [Thu, 8 Jun 2023 08:57:02 +0000 (08:57 +0000)]
add tests related to --libs-only-l and the likes

remove extraneous spaces from reference output now that we're closer
to the original pkg-config

16 months agomove to perl use v5.36
espie [Thu, 8 Jun 2023 08:55:27 +0000 (08:55 +0000)]
move to perl use v5.36
also fix a discrepancy wrt the "original" pkg-config

thanks to tb@ for testing.

17 months agoRename ifconfig tcprecvoffload to tcplro. It is shorter and
bluhm [Wed, 7 Jun 2023 18:42:40 +0000 (18:42 +0000)]
Rename ifconfig tcprecvoffload to tcplro.  It is shorter and
more consistent.
discussed with jan@ mvs@ chris@ claudio@ dlg@

17 months agoDocument CMS signing-time <> mod-time trick
job [Wed, 7 Jun 2023 16:23:02 +0000 (16:23 +0000)]
Document CMS signing-time <> mod-time trick

17 months agodo not pass @_ to code snippets, prepare for 5.36
espie [Wed, 7 Jun 2023 15:09:01 +0000 (15:09 +0000)]
do not pass @_ to code snippets, prepare for 5.36

found out by aja@

17 months agoAdd portable version and m88k-specific version lb() function, because
aoyama [Wed, 7 Jun 2023 12:56:22 +0000 (12:56 +0000)]
Add portable version and m88k-specific version lb() function, because
unfortunately gcc3 does not have __builtin_clz().

ok miod@ otto@

17 months agoCosmetic tweak for previous
tb [Wed, 7 Jun 2023 11:09:08 +0000 (11:09 +0000)]
Cosmetic tweak for previous

Once we expect ASPA version 1 and someone sends us version 0, make that
explicit instead of complaining about ASN1_INTEGER_get_uint64() failing.

ok job

17 months agoRefer to the field "thisUpdate" instead of the non-existent "lastUpdate".
schwarze [Wed, 7 Jun 2023 10:53:30 +0000 (10:53 +0000)]
Refer to the field "thisUpdate" instead of the non-existent "lastUpdate".
Similar to X509_get0_notBefore(3) rev. 1.6.
Requested by and OK tb@.

17 months agoIn anticipation of a bump of the ASPA eContent profile version, update
job [Wed, 7 Jun 2023 10:46:34 +0000 (10:46 +0000)]
In anticipation of a bump of the ASPA eContent profile version, update
valid_econtent_version() to allow for non-zero versions.

OK tb@

17 months agoupon resume, fpureset() was being called prematurely (before cpu_init,
deraadt [Wed, 7 Jun 2023 04:46:09 +0000 (04:46 +0000)]
upon resume, fpureset() was being called prematurely (before cpu_init,
which does not matter today, but will matter a lot in near future).
But actually it isn't needed at all, cpu_init() does it again.
So remove the call.
ok guenther

17 months agoIn 1995, Eric A. Young chose a confusing name for the "lastUpdate" field
schwarze [Tue, 6 Jun 2023 16:20:13 +0000 (16:20 +0000)]
In 1995, Eric A. Young chose a confusing name for the "lastUpdate" field
of the X509_CRL_INFO object.  It should have been called "thisUpdate"
like in RFC 5280 section 5.1 (and in its precursor RFC 2459).  Then again,
RFC 2459 was only published in 1999, so maybe the terminology wasn't
firmly established yet when Young wrote his code several years earlier -
just guessing, neither we nor the OpenSSL folks appear to know the real
reasons...

Anyway, we have been stuck with the "lastUpdate" names in the API for
more than two decades now, so clarify in the documentation what they
refer to and what they really mean.

Requested by and OK tb@.

17 months agoFix typo in comment: exta -> extra
tb [Tue, 6 Jun 2023 16:10:56 +0000 (16:10 +0000)]
Fix typo in comment: exta -> extra

17 months agoUse same pattern to work with offset by using a uint8_t pointer that
claudio [Tue, 6 Jun 2023 16:09:35 +0000 (16:09 +0000)]
Use same pattern to work with offset by using a uint8_t pointer that
is loaded with ibuf_data(). This is by no means better but allows to
switch ibuf_data() to return void *.
OK tb@

17 months agoMake the tlsv1.0 and tlsv1.1 options in relayd do nothing
beck [Tue, 6 Jun 2023 15:16:52 +0000 (15:16 +0000)]
Make the tlsv1.0 and tlsv1.1 options in relayd do nothing

Also document that fact, and that the existing ssl3 option
does nothing. This changes relayd to no longer request tls1.0
or tls1.1 in preparation for the upcoming deprecation of these
out of data protocols

ok jsing@ bluhm@ tb@ claudio@ benno@

17 months agoUse ibuf_seek() instead of ibuf_data() + offset constructs. Effect is
claudio [Tue, 6 Jun 2023 13:27:49 +0000 (13:27 +0000)]
Use ibuf_seek() instead of ibuf_data() + offset constructs. Effect is
the same in these cases.
OK tb@

17 months agoFold disk crypto question '?' text into prompt
kn [Tue, 6 Jun 2023 09:35:44 +0000 (09:35 +0000)]
Fold disk crypto question '?' text into prompt

Since this question moved after the root disk one and '?' stopped listing disks,
hoist the implementation details in order to drop the custom answer and reuse
existing ask_yn(), thus

Encrypt the root disk? (yes, no or '?' for details) [no] ?
Create a passphrase protected CRYPTO softraid volume to be used as root disk.
Encrypt the root disk? (yes, no or '?' for details) [no]
becomes
Encrypt the root disk? (passphrase CRYPTO softraid) [no]

Prodded by afresh1

17 months agodon't need mcx_uptime() now that we have nsecuptime()
dlg [Tue, 6 Jun 2023 01:40:04 +0000 (01:40 +0000)]
don't need mcx_uptime() now that we have nsecuptime()

ok jmatthew@

17 months agoFix copy+paste error in x509 asn regress
job [Mon, 5 Jun 2023 18:32:06 +0000 (18:32 +0000)]
Fix copy+paste error in x509 asn regress

17 months agoImprove the description of CMS_get0_signers()
job [Mon, 5 Jun 2023 17:17:23 +0000 (17:17 +0000)]
Improve the description of CMS_get0_signers()

Suggestion from Małgorzata Olszówka, they noted:

"The original wording suggests that it is required to execute
CMS_get0_signers() after CMS_verify(), while it is CMS_get0_signers()
that requires prior successful invocation of CMS_verify()."

OK tb@

17 months agoSync json.c with rpki-client rev 1.3:
claudio [Mon, 5 Jun 2023 16:24:05 +0000 (16:24 +0000)]
Sync json.c with rpki-client rev 1.3:
Add an extra argument compact to json_do_object() to instruct the parser
to dump this object on a single line.

While one can select on an object to object basis for arrays the compact
setting is inherited from the surrounding object.

OK tb@