IABSD.fr/src/sys
Branch :
Log
| Author | Commit | Date | CI | Message |
|---|---|---|---|---|
 |
3dc5ebc3 | 2025-02-25 22:10:39 | sys/utvfu, sys/uvideo: clearify used timestamps OK: mglocker@ | |
 |
16c6aa4a | 2025-02-25 11:29:17 | Add new counters for lock upgrades. While here unify re-lock counters to reduce the number of atomic operations. ok tb@ | |
|
924e3373 | 2025-02-24 21:24:54 | sys/uvideo: never try to allocate more than MALLOC_MAX OK: miod@ | |
 |
0f757f3d | 2025-02-24 20:16:14 | IPsec path MTU uses routing table before pf switches it. If pf(4) switches the rtable, the route for path MTU discovery must be generated in the original routing table. For that ip_output() keeps the original rtableid. Then a local TCP socket uses the correct route. This did not work when IPsec was involed. Pass orig_rtableid also to ip_output_ipsec_send() to use the same logic in ip_output_ipsec_pmtu_update(). A similar change is necessary for ip6_output() and ip6_forward(). OK markus@ | |
 |
c352d04b | 2025-02-24 18:07:50 | Use mallocarray() rather than malloc(a * b) This can't overflow but it smells bad. ok deraadt | |
 |
07477e78 | 2025-02-24 15:43:55 | It is unlikely that pinsyscalls() will ever be unlocked, so remove the comment proposing it. ok kettenis | |
|
b4359f85 | 2025-02-24 15:42:35 | Only allow pinsyscalls() to set the libc pins once. Mateusz Guzik identified a race where after the copyin sleeping point, memory would be leaked. There are various solutions, but choosing one worked on with kettenis. | |
 |
1e9ab3e3 | 2025-02-24 13:18:01 | reguster -> register | |
|
183cafad | 2025-02-24 12:43:29 | sys/uvideo: dump bDescriptorSubtype for unsupported CS_INTERFACE when debug is enabled OK: mglocker@ | |
 |
49f05fab | 2025-02-24 09:40:01 | Refactor LRO turn off code Its easier to turn off LRO via ioctl calls inside of several hardware and pseudo interfaces. Thus, we avoid manipulating internal data structures form the outside and avoid unnecessary reinitializations. Tested by bluhm@ OK bluhm@ | |
|
eda3d7dc | 2025-02-24 03:29:12 | drm/amdgpu: avoid buffer overflow attach in smu_sys_set_pp_table() From Jiang Liu 2498d2db1d35e88a2060ea191ae75dce853dd084 in linux-6.12.y/6.12.16 1abb2648698bf10783d2236a6b4a7ca5e8021699 in mainline linux | |
|
1b92bee2 | 2025-02-24 03:27:14 | drm: Fix DSC BPP increment decoding From Imre Deak 740bc21e25825c67444cd94ecd6b08669d48e95c in linux-6.12.y/6.12.16 e00a2e5d485faf53c7a24b9d1b575a642227947f in mainline linux | |
|
98999899 | 2025-02-24 03:20:25 | drm/amdgpu: bail out when failed to load fw in psp_init_cap_microcode() From Jiang Liu d1d10bd595539ed82ab59b60249f9bdf0994f678 in linux-6.12.y/6.12.16 a0a455b4bc7483ad60e8b8a50330c1e05bb7bfcf in mainline linux | |
|
21380ad5 | 2025-02-24 03:17:30 | amdkfd: properly free gang_ctx_bo when failed to init user queue From Zhu Lingshan ae5ab1c1ae504f622cc1ff48830a9ed48428146d in linux-6.12.y/6.12.16 a33f7f9660705fb2ecf3467b2c48965564f392ce in mainline linux | |
|
e24267e0 | 2025-02-24 03:13:56 | drm/i915/selftests: avoid using uninitialized context From Krzysztof Karas 19a6d7c28278247e2433cfd7860fb0ea36aee2cd in linux-6.12.y/6.12.16 53139b3f9998ea07289e7b70b909fea2264a0de9 in mainline linux | |
|
ed1f1a00 | 2025-02-24 00:50:11 | drm/i915/dp_mst: Fix connector initialization in intel_dp_add_mst_connector() From Imre Deak d58f65df2dcb59acd3965907507cfa608fe924b4 in mainline linux avoids a NULL deref when using a thunderbolt dock, reported by lorenz on bugs@ Johnny Epsom reported the same problem with a usb-c dock on misc@ | |
|
23fac5c5 | 2025-02-24 00:19:59 | add comments for more revisions of known ids 747e rev d8 is Radeon RX 7800M 7480 rev c2 is Radeon RX 7650 GRE found in AMD Software: Adrenalin Edition 25.2.2 | |
|
5f02adc7 | 2025-02-23 14:12:15 | sys/uvideo: compute max_packet_size the same way in all cases OK: mglocker@ | |
|
a5c838d8 | 2025-02-23 08:28:57 | sys/uvideo: dump baInterfaceNr when debug is enabled OK: mglocker@ | |
|
45a54130 | 2025-02-21 22:21:20 | Move kassert from resolve to add case in rtrequest(). In case RTM_RESOLVE there is already an assertion about ifa_ifp != NULL. Move it down after the fallthrough to cover also RTM_ADD. This should give a better hint from syzkaller what is going wrong. Reported-by: syzbot+f77fe03091e5efd9aaf9@syzkaller.appspotmail.com OK claudio@ | |
 |
9a71ea36 | 2025-02-21 06:20:12 | replace "if (!task_del) taskq_barrier" with "taskq_del_barrier". as per src/sys/kern/kern_task.c r1.36, it's possible for a task to be re-added while it's currently running. in this situation the "if (!task_del)" skips the barrier but doesn't do anything about the currently running code, which taskq_del_barrier properly handles. ---------------------------------------------------------------------- | |
|
cf9e9cef | 2025-02-21 03:34:10 | use BACKLIGHT_POWER defines for power properties; no functional change | |
|
d8ec5edc | 2025-02-20 11:18:42 | remove uneeded ifdef | |
|
e17cfef6 | 2025-02-20 02:04:42 | rename inteldrm_softc to drm_i915_private, reduces diff to linux | |
 |
8f64efa5 | 2025-02-19 21:33:10 | Add efi(4) such that we can use installboot -c in the installer. ok deraadt@, kn@ | |
 |
483a78e1 | 2025-02-19 18:25:03 | Add some more PPC_FEATURES2 defines From FreeBSD via Brad, ok kettenis@ | |
|
1e4b7a50 | 2025-02-19 12:15:09 | sync with upstream This replaces two #ifdef FOO with #if FOO-0 for unistd.h and stdarg.h, which is mostly a noop. Base carries local patches to replace these with #if 1, so it won't have this. I should revisit this at some point. | |
|
9a32430e | 2025-02-19 12:13:33 | sync with base | |
|
12a8400b | 2025-02-19 11:10:54 | Nullify `uanon' and `uobject' pointer while holding the corresponding lock. Stop modifying these pointers inside the pmemrange allocator. Assert instead that physical page descriptors are no longer associated to objects when they are given back to the allocator. ok miod@ | |
|
a4a5959e | 2025-02-19 11:07:47 | Document vm_page locking. | |
|
85b55483 | 2025-02-19 10:38:56 | regen | |
|
582d7a8e | 2025-02-19 10:38:14 | add AMD "Krackan Point" APU device id 1114 rev c2 is Radeon 860M 1114 rev c3 is Radeon 840M 1114 rev d2 is Radeon 860M 1114 rev d3 is Radeon 840M found in Asus Vivobook S16 (M5606KA) graphics driver | |
|
02f8071d | 2025-02-19 08:45:27 | regen | |
|
f8ccd129 | 2025-02-19 08:44:56 | sort | |
|
47512c81 | 2025-02-19 02:16:24 | regen | |
|
b9b3036b | 2025-02-19 02:15:33 | add AMD "Strix Halo" APU device id, and another Navi 31 1586 rev c1 is Radeon 8060S 1586 rev c2 is Radeon 8050S 1586 rev d1 is Radeon 8060S 1586 rev d2 is Radeon 8050S found in Asus ROG Flow Z13 (2025) graphics driver 7449 rev 00 is Radeon Pro W7800 48GB found in AMD Software: Adrenalin Edition 25.2.1 | |
|
181c8b93 | 2025-02-18 09:18:57 | Use the same system register alias trick as we have in cpu.c to make clang19 happy. ok patrick@ | |
|
1ba4a5b9 | 2025-02-18 05:38:09 | reduce diff to linux | |
|
a11201ac | 2025-02-17 22:31:53 | Add support for device mappings in pmap_kenter_pa(9) the same way we handle those in pmap_enter(9). arm64 diff by kettenis@, ok kettenis@ | |
|
ed913ab5 | 2025-02-17 21:08:40 | Fix pre-registration of wakeup interrupts. ok patrick@ | |
|
4039bfa0 | 2025-02-17 20:31:25 | Handle RTF_GATEWAY route with rt_gwroute NULL. rtrequest_delete() calls rt_putgwroute() to set rt_gwroute to NULL. When another thread holds a reference to such a route, an assertion failed in rtisvalid() and rt_getll(). Handle this case, rt_getll() may return NULL then. OK claudio@ | |
 |
f7a1cf18 | 2025-02-17 15:45:55 | Introduce ps_trapped -- a pointer to the struct proc that is stopped in the debugger trap. In proc_trap set and unset ps_trapped together with ps_xsig right before and after suspending. In ptrace(2) use this information to have process_tprfind() always return that thread and make all of that code no longer depend on ps_single. Only access ps_trapped holding the ps_mtx to ensure the data is valid. Feedback and OK kettenis@ | |
|
c05afcd0 | 2025-02-17 13:28:26 | Disable pagefault handling when unwinding userland stack. When collecting userland stacktraces with dt(4), the probe may run in an interrupt routine at IPL_VM or higher. Resolving pagefaults in this case might result in panic or corrupt UVM data structures. Userland stack frames cannot be trusted, a bogus frame pointer may force us into a page fault. From Christian Ludwig, ok kettenis@ | |
|
d4b87b93 | 2025-02-17 13:10:56 | regen | |
|
4541be0b | 2025-02-17 13:10:27 | Push KERNEL_LOCK() inside __realpath(2). ok semarie@, tb@ | |
|
7a23ec91 | 2025-02-17 12:46:02 | Toeplitz hash for UDP and IPv6 TCP output. IPv4 TCP output uses the toeplitz hash as flow id. It is calculated in in_pcbconnect() and can be used for all connected sockets. Add it for UDP and TCP IPv6, too. As pf calculates its own hash, this affects only setups with pf disabled. It gives an improvement in traffic distribution over the queues and 20% performance increase with UDP send on v4/v6 and TCP send on v6 without pf. tested and OK sf@ | |
|
257f1919 | 2025-02-17 12:16:20 | Revert "drm/amd/display: Fix green screen issue after suspend" From Rodrigo Siqueira c963ef1840a09cc19e52cce3f2f9d962a2a5092d in linux-6.12.y/6.12.14 04d6273faed083e619fc39a738ab0372b6a4db20 in mainline linux | |
|
7ff55e83 | 2025-02-17 12:13:03 | drm/amd/display: Fix seamless boot sequence From Lo-an Chen 7173b70df472813095bedce5798a1cb79c0ba579 in linux-6.12.y/6.12.14 e01f07cb92513ca4b9b219ab9caa34d607bc1e2d in mainline linux | |
|
d697aa04 | 2025-02-17 12:09:42 | drm/amdgpu: add a BO metadata flag to disable write compression for Vulkan From Marek Olsak 20a57f68db1b7e8de97acd1f04bb697f3c6ad589 in linux-6.12.y/6.12.14 2255b40cacc2e5ef1b127770fc1808c60de4a2fc in mainline linux | |
|
d6f42f09 | 2025-02-17 12:05:55 | drm/i915: Drop 64bpp YUV formats from ICL+ SDR planes From Ville Syrjala 4b1d3186284ff015b90fa838e663e300ab163153 in linux-6.12.y/6.12.14 c7b49506b3ba7a62335e6f666a43f67d5cd9fd1e in mainline linux | |
|
6791f72f | 2025-02-17 12:03:34 | drm/i915/dp: Iterate DSC BPP from high to low on all platforms From Jani Nikula c842f22fc26eef56cc3f594f18f8af5d36e0ffa7 in linux-6.12.y/6.12.14 230b19bc2bcc5897d0e20b4ce7e9790a469a2db0 in mainline linux | |
|
7dfd7741 | 2025-02-17 12:01:28 | drm/i915: Fix page cleanup on DMA remap failure From Brian Geffon 2708a151e0e9943e90a36ebec36103586c41c424 in linux-6.12.y/6.12.14 fa6182c8b13ebfdc70ebdc09161a70dd8131f3b1 in mainline linux | |
|
d93ced68 | 2025-02-17 11:59:27 | drm/i915/guc: Debug print LRC state entries only if the context is pinned From Daniele Ceraolo Spurio 5775ef49d792e55756aa9cfbad684fc887fb1520 in linux-6.12.y/6.12.14 57965269896313e1629a518d3971ad55f599b792 in mainline linux | |
|
693a9725 | 2025-02-17 11:57:32 | Revert "drm/amd/display: Use HW lock mgr for PSR1" From Tom Chung 95c75578c420110c43791295985abb961d6dc033 in linux-6.12.y/6.12.14 f245b400a223a71d6d5f4c72a2cb9b573a7fc2b6 in mainline linux | |
|
2bc1d348 | 2025-02-17 11:55:38 | drm/amdkfd: Block per-queue reset when halt_if_hws_hang=1 From Jay Cornwall 4d23b42a45dcff725dc8be8c41aa4712241dd764 in linux-6.12.y/6.12.14 f214b7beb00621b983e67ce97477afc3ab4b38f4 in mainline linux | |
|
d7d82365 | 2025-02-17 11:54:05 | drm/amdkfd: only flush the validate MES contex From Prike Liang 89dcc75d517208bae58ac15fd97a56721a529ade in linux-6.12.y/6.12.14 9078a5bfa21e78ae68b6d7c365d1b92f26720c55 in mainline linux | |
|
4e63db5a | 2025-02-17 11:52:37 | drm/amd/amdgpu: change the config of cgcg on gfx12 From Kenneth Feng 7e207a094ee69b2368a4c1d11c1863f7134945a4 in linux-6.12.y/6.12.14 5cda56bd86c455341087dca29c65dc7c87f84340 in mainline linux | |
|
9af075f5 | 2025-02-17 11:51:24 | drm/amd/pm: Mark MM activity as unsupported From Lijo Lazar 721a493fa2a5e9937f3cf4069a9dd3b423b19d5a in linux-6.12.y/6.12.14 819bf6662b93a5a8b0c396d2c7e7fab6264c9808 in mainline linux | |
|
7035daaf | 2025-02-17 11:50:06 | drm/amd/display: Optimize cursor position updates From Aric Cyr 3ec8e3dab60c3e556bf542f9cda40d9d54b12c7b in linux-6.12.y/6.12.14 024771f3fb75dc817e9429d5763f1a6eb84b6f21 in mainline linux | |
|
824a42de | 2025-02-17 11:48:00 | drm/client: Handle tiled displays better From Maarten Lankhorst ac7b5f3e4d5beeb0e59ec75c4a6055110a46fb29 in linux-6.12.y/6.12.14 10026f536843eb8c9148ef6ffb4c6deeebc26838 in mainline linux | |
|
5c428613 | 2025-02-17 11:45:36 | drm/modeset: Handle tiled displays in pan_display_atomic. From Maarten Lankhorst b865c6a6178b0828aca62b8fe736f24f04d912f3 in linux-6.12.y/6.12.14 f4a9dd57e549a17a7dac1c1defec26abd7e5c2d4 in mainline linux | |
|
287a0281 | 2025-02-17 11:43:45 | drm/i915/dp: fix the Adaptive sync Operation mode for SDP From Ankit Nautiyal 4c6ed049bc4ed907345dc6aa3a33e066569e0853 in linux-6.12.y/6.12.14 4466302262b38f5e6c65325035b4036a42efc934 in mainline linux | |
|
f12786ce | 2025-02-17 11:42:03 | drm/amdgpu: Fix Circular Locking Dependency in AMDGPU GFX Isolation From Srinivasan Shanmugam 4e144498d16baa1be26e7ba8837483697a583900 in linux-6.12.y/6.12.14 1e8c193f8ca7ab7dff4f4747b45a55dca23c00f4 in mainline linux | |
|
e8ee1e2e | 2025-02-17 11:39:53 | drm/amd/display: Limit Scaling Ratio on DCN3.01 From Gabe Teeger d1544dc32c67c80b9e3512fa4e187931be1a9e8c in linux-6.12.y/6.12.14 abc0ad6d08440761b199988c329ad7ac83f41c9b in mainline linux | |
|
b47c1b83 | 2025-02-17 11:38:13 | drm/amdkfd: Queue interrupt work to different CPU From Philip Yang d76dc02531a610b3f7c2e3262b14137137787e0e in linux-6.12.y/6.12.14 34db5a32617d102e8042151bb87590e43c97132e in mainline linux | |
|
70ae38f9 | 2025-02-17 11:36:52 | drm/amdgpu: Don't enable sdma 4.4.5 CTXEMPTY interrupt From Philip Yang 9f952450337502ca3f20c29599b8bfb14327e6ec in linux-6.12.y/6.12.14 b4b7271e5ca95b581f2fcc4ae852c4079215e92d in mainline linux | |
|
fdc47235 | 2025-02-17 11:35:02 | drm/amd/display: Fix Mode Cutoff in DSC Passthrough to DP2.1 Monitor From Fangzhi Zuo b79eaf9b86a8789f15a2ed05c6ffe2b4e5e35aea in linux-6.12.y/6.12.14 e56ad45e991128bf4db160b75a1d9f647a341d8f in mainline linux | |
|
cc0343a4 | 2025-02-17 11:31:01 | drm/radeon: use eld_mutex to protect access to connector->eld From Dmitry Baryshkov 1cd3e56a07acd34cc389d490a259ad2a1dd6f282 in linux-6.12.y/6.12.14 b54c14f82428c8a602392d4cae1958a71a578132 in mainline linux | |
|
6b94d9b3 | 2025-02-17 11:28:23 | drm/amd/display: use eld_mutex to protect access to connector->eld From Dmitry Baryshkov cd604e8e993cd6f6e8aab0a8d23052885de60472 in linux-6.12.y/6.12.14 819bee01eea06282d7bda17d46caf29cae4f6d84 in mainline linux | |
|
04766a0e | 2025-02-17 11:26:39 | drm/connector: add mutex to protect ELD from concurrent access From Dmitry Baryshkov b006aadf72f933770a6d66bc0985c77417f6b4e4 in linux-6.12.y/6.12.14 df7c8e3dde37a9d81c0613285b43600f3cc70f34 in mainline linux | |
|
107bc997 | 2025-02-17 11:21:59 | drm/amd/display: Overwriting dualDPP UBF values before usage From Ausef Yousof 0412bca906d598b0a60c0da837dadb036a95db84 in linux-6.12.y/6.12.14 24909d9ec7c3afa8da2f3c9afa312e7a4a61f250 in mainline linux | |
|
e71392c7 | 2025-02-17 11:19:51 | drm/amd/display: Populate chroma prefetch parameters, DET buffer fix From Ausef Yousof d25d5eb35905ade48b2ca657f30ea9b2e5804141 in linux-6.12.y/6.12.14 70fec46519fca859aa209f5f02e7e0a0123aca4a in mainline linux | |
|
d2d65dd8 | 2025-02-17 11:16:52 | drm: Add panel backlight quirks From Thomas Weissschuh c5842f1dade0c2899e14f2f9d7de7103dbf1d678 in linux-6.12.y/6.12.14 22e5c7ae12145af13785e3ff138395d5b1a22116 in mainline linux | |
|
a10b347c | 2025-02-17 10:16:05 | Use ps_mtx to lock the child process that is being checked by dowait6. The checks of the various ps_flags field is all but atomic so we need to lock the process before doing all these checks. Only that way a reliable result is seen. It also ensures that the child process did finish the transition before signaling up to the parent. | |
|
9151e582 | 2025-02-17 10:07:10 | Make wakeup of parent process in dowait6 reliable even without kernel lock. Currently dowait6 uses tsleep to wait for an event to report back to wait4(2). This only works correctly if all wakeups also use the kernel lock. Instead of this use an atomic flag to indicate if a process has pending wait events and use sleep_setup()/sleep_finish() to reliably enter sleep without losing a wakeup. dowait6 clears the PS_WAITEVENT flag at the start of its loop over all child processes and if at the end of the loop the flag is still unset the thread goes to sleep. Before all wakeups the PS_WAITEVENT flag is now set so that dowait6 knows that more work is pending and restarts the scan loop. This is needed since more and more code is unlocked and we end up missing wakeups which results in hanging processes. | |
 |
2f5a0aea | 2025-02-17 08:56:33 | Get rid of unused `so' argument in sbreserve(). ok bluhm | |
|
edfe48e7 | 2025-02-17 08:39:50 | Reorder <sys/socketvar.h> by moving inline functions to the end. ok bluhm | |
|
4df8ba73 | 2025-02-17 05:00:06 | reduce diff to linux | |
|
ae0be371 | 2025-02-16 16:05:07 | Use shared socket lock for NFS. The socket functions that are called by NFS code are MP safe. Shared net lock together with socket lock is sufficient. OK mvs@ | |
 |
c48feb94 | 2025-02-16 13:07:08 | Argument order fix for m_gethdr ok martijn@ claudio@ bluhm@ | |
|
3411e7e2 | 2025-02-16 11:39:28 | Revert SMR protection of rt_gwroute. Using a smr_barrier() in rt_putgwroute() slows down adding routes. This is the hot path for BGP router. Syncing the FIB is now taking ages and the system is close to unrespnsive in that time. found by claudio@ | |
 |
c129e66c | 2025-02-16 10:11:37 | Add support for reading eeprom pages for the few cards with SFP slots. ok dlg@ | |
|
96ec60f5 | 2025-02-15 21:49:28 | Fail hard with mutex assertion if socket buffer is not locked. OK mvs@ | |
|
a1e112ca | 2025-02-15 11:55:17 | call intel_fbdev_invalidate() at the end of intelfb_create() Avoids a problem seen with the (Alder Lake) Framework 13 after the 6.12 drm update. Userland console output did not display without switching to a virtual terminal and back. Not all Alder Lake machines, T14 Gen 3 Intel worked as expected before this change. reported and tested by matthieu@ | |
|
5c4be1f5 | 2025-02-15 09:05:15 | sys/uvideo: clarify bulk endpoint stream closure UVC does not specify how to notify a bulk-based device when the video stream stops. Both Linux and Windows send a CLEAR_FEATURE(HALT) request to the video streaming bulk endpoint. OK: mglocker@ | |
|
8d246409 | 2025-02-14 18:42:43 | It is important that we map memory with the right attributes otherwise we may trigger an SError. The crashlog entries for the SMC live in SRAM which needs to be mapped as device memory. To make this possible, make the rk_logmap() rtkit callback function return the desired memory attributes in the low bits of the address, using PMAP_NOCACHE or PMAP_DEVICE defines. Rewrite the DCP implementation of rk_logmap() to do a better job ati determining the correct mapping. The current code would fail in some cases (i.e. on the M1 Pro MacBook Pro) and map some unused address as normal non-cachable which could also trigger an SError. ok jca@ | |
|
4a7354f2 | 2025-02-14 18:36:04 | Add support for device mappings in pmap_kenter_pa(9) the same way we handle those in pmap_enter(9). ok jca@ | |
 |
e8902928 | 2025-02-14 13:29:00 | Add the kern.audio.kbdcontrol sysctl(2) variable If set to 0, the volume keys on multimedia keyboards are handled as regular keys, allowing programs to use them. The default is 1, i.e. the current behavior. ok armani, kettenis, mvs | |
|
51442b8a | 2025-02-14 13:14:13 | add tunneldf support to sec(4) sec(4) is a very thin wrapper around the existing ipsec output processing for encapsulating packets, and inherited the behaviour that the DF flag was propagated from the encapsulated packet to the outer ip header. this means if the sec(4) interface has a large mtu and is carrying packets with DF set over a network that can't transport large(r) packets, these packets are effectively dropped. ipsec applied via the SPD copes with this by having SAs figure out the path mtu and using that when applying policy, but sec(4) is an interface, so the network stack uses the interface mtu rather than the associated SA path mtu. rfc4459 discusses this kind of problem has offers a variety of solutions. this implements one of the simpler options, which is to allow the tunnel endpoints to manage the DF regardless of the payload and reassemble the encapsulated packets. to actually do this, ipsec output packet processing has to be able to take an argument that says how you want DF to be handled. in the future we're going to look at how we can use the path mtu determined by the ipsec SA to try and implement one of the other solutions from the RFC, which is to signal the lower mtu to the sources of tunnelled packets. tested by and ok claudio@ | |
 |
c33cf9b0 | 2025-02-14 04:56:26 | Enable mtrng(4). ok kettenis@ | |
|
e04b0fdd | 2025-02-14 03:11:05 | Add mtrng(4), supporting the 32-bit random number generator on MediaTek SoCs. ok kettenis@ | |
|
dddbedba | 2025-02-13 21:01:34 | Fix route entry race when accessing rt_gwroute. Kassert in rt_getll() was triggered as rt_gwroute could be NULL. Problem was introduced by shared netlock around tcp_timer_rexmt(). PMTU discovery calls rtrequest_delete() which was missing proper locking around rt_gwroute. As rt_getll() is called by ARP and ND6 resolve in the hot path, use SMR to provide the pointer to rt_gwroute lockless. Reference count of the returned route is incremented, caller has to free it. Modifying rt_gwroute or rt_cachecnt in rt_putgwroute() is protected by per route lock. OK mvs@ | |
 |
88a67239 | 2025-02-13 19:54:44 | Remove leftover code from a previous iteration of this and which should have been removed before committing this in the first place. | |
|
9470f526 | 2025-02-13 14:44:33 | Get rid of unused `so' argument in sbrelease(). No functional changes. ok bluhm | |
|
babf29f6 | 2025-02-13 12:39:15 | Remove socket lock from soo_ioctl(FIOASYNC). The socket buffer flags are protected by socket buffer mutex. Neither socket lock nor net lock are needed anymore. OK mvs@ | |
|
44a9493f | 2025-02-12 21:41:53 | Use shared net lock in sysctl fill file. The macro FILLINPTABLE() already uses shared net lock for fill_file(). The FILLIT() sysctls should do the same instead of grabbing exclusive net lock. OK mvs@ | |
|
bd92615b | 2025-02-12 21:28:10 | Use socket lock for inpcb notify. The notify and ctlinput functions were not MP safe. They need socket lock which can be aquired by in_pcbsolock_ref(). Of course in_pcbnotifyall() has to be called without holding a socket lock. Rename in_rtchange() to in_pcbrtchange(). This is the correct namespace and the functions take care of the inpcb route. OK mvs@ | |
|
eb8501b6 | 2025-02-12 20:18:31 | Fix ddb machine sysregs stack overflow. In 64-bit mode, the operand size of SIDT and SGDT instructions is 8+2 bytes. On amd64 ddb command 'machine sysregs' reserved only int64_t, resulting in a stack overflow. Use struct region_descriptor which has the correct memory layout for IDTR and GDTR. Allow to call db_sysregs_cmd() also with single processor kernel. from Sebastian Sturm; OK deraadt@ miod@ | |
|
99f2f888 | 2025-02-12 14:11:26 | Changes to the per-process unveil datastructures can be raced by either pledge() [removing all path promises] or unveil() [adding new paths], against namei() inspecting in other thread system calls. Detailed locking for this would make future work on vfs unlocking difficult, so use the simplest way for now: SINGLE_UNWIND ok claudio | |
|
705b70e0 | 2025-02-11 22:27:09 | Implement support for SVE (Scalable Vector Extension) on arm64. ok patrick@ |