| Message ID | 20221106150243.150437-1-Jason@zx2c4.com |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [v3] hw_random: use add_hwgenerator_randomness() for early entropy | expand |
On Sun, Nov 06, 2022 at 04:02:43PM +0100, Jason A. Donenfeld wrote: > Rather than calling add_device_randomness(), the add_early_randomness() > function should use add_hwgenerator_randomness(), so that the early > entropy can be potentially credited, which allows for the RNG to > initialize earlier without having to wait for the kthread to come up. > > This requires some minor API refactoring, by adding a `sleep_after` > parameter to add_hwgenerator_randomness(), so that we don't hit a > blocking sleep from add_early_randomness(). > > Cc: Herbert Xu <herbert@gondor.apana.org.au> > Cc: Dominik Brodowski <linux@dominikbrodowski.net> > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > --- > Herbert - it might be easiest for me to take this patch if you want? Or > if this will interfere with what you have going on, you can take it. Let > me know what you feel like. -Jason I don't have anything that touches this file so feel free to push it through your tree: Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Cheers,
Hi Herbert, On Mon, Nov 7, 2022 at 3:17 AM Herbert Xu <herbert@gondor.apana.org.au> wrote: > > On Sun, Nov 06, 2022 at 04:02:43PM +0100, Jason A. Donenfeld wrote: > > Rather than calling add_device_randomness(), the add_early_randomness() > > function should use add_hwgenerator_randomness(), so that the early > > entropy can be potentially credited, which allows for the RNG to > > initialize earlier without having to wait for the kthread to come up. > > > > This requires some minor API refactoring, by adding a `sleep_after` > > parameter to add_hwgenerator_randomness(), so that we don't hit a > > blocking sleep from add_early_randomness(). > > > > Cc: Herbert Xu <herbert@gondor.apana.org.au> > > Cc: Dominik Brodowski <linux@dominikbrodowski.net> > > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > > --- > > Herbert - it might be easiest for me to take this patch if you want? Or > > if this will interfere with what you have going on, you can take it. Let > > me know what you feel like. -Jason > > I don't have anything that touches this file so feel free to push > it through your tree: > > Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Okay, will do. But by the way, feel free to change your mind about this if need be. For example, I sent another patch that touches core.c too (the entropy quality one). It touches different lines, so there shouldn't be a conflict, but if it's still annoying for you and you want to take them both, just pipe up and I'll drop this one from my tree. Jason
On Mon, Nov 07, 2022 at 01:55:53PM +0100, Jason A. Donenfeld wrote: > > Okay, will do. But by the way, feel free to change your mind about > this if need be. For example, I sent another patch that touches core.c > too (the entropy quality one). It touches different lines, so there > shouldn't be a conflict, but if it's still annoying for you and you > want to take them both, just pipe up and I'll drop this one from my > tree. They look like they shouldn't really conflict so it should be fine. Thanks,
Il 06/11/22 16:02, Jason A. Donenfeld ha scritto: > Rather than calling add_device_randomness(), the add_early_randomness() > function should use add_hwgenerator_randomness(), so that the early > entropy can be potentially credited, which allows for the RNG to > initialize earlier without having to wait for the kthread to come up. > > This requires some minor API refactoring, by adding a `sleep_after` > parameter to add_hwgenerator_randomness(), so that we don't hit a > blocking sleep from add_early_randomness(). > > Cc: Herbert Xu <herbert@gondor.apana.org.au> > Cc: Dominik Brodowski <linux@dominikbrodowski.net> > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > Acked-by: Herbert Xu <herbert@gondor.apana.org.au> > --- > Herbert - it might be easiest for me to take this patch if you want? Or > if this will interfere with what you have going on, you can take it. Let > me know what you feel like. -Jason > > drivers/char/hw_random/core.c | 8 +++++--- > drivers/char/random.c | 12 ++++++------ > include/linux/random.h | 2 +- > 3 files changed, 12 insertions(+), 10 deletions(-) > Hello, I tried booting next-20221108 on Acer Tomato Chromebook (MediaTek MT8195) but this commit is producing a kernel panic. Trace follows: [ 2.957452] cr50_i2c 3-0050: cr50 TPM 2.0 (i2c 0x50 irq 114 id 0x28) [ 3.047508] ------------[ cut here ]------------ [ 3.052178] WARNING: CPU: 5 PID: 1 at kernel/kthread.c:75 kthread_should_stop+0x2c/0x44 [ 3.060270] Modules linked in: [ 3.063359] CPU: 5 PID: 1 Comm: swapper/0 Not tainted 6.1.0-rc4-next-20221108+ #111 [ 3.071086] Hardware name: Acer Tomato (rev2) board (DT) [ 3.076447] pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 3.083473] pc : kthread_should_stop+0x2c/0x44 [ 3.087962] lr : add_hwgenerator_randomness+0x6c/0x110 [ 3.093152] sp : ffff80000808b570 [ 3.096499] x29: ffff80000808b570 x28: 0000000000061a80 x27: 00000000000f4240 [ 3.103707] x26: ffff030f3efe9d78 x25: ffffd855b3930000 x24: 0000000000000000 [ 3.110915] x23: 0000000000000000 x22: ffffd855b2c00d08 x21: 0000000000000020 [ 3.118122] x20: 0000000000000000 x19: ffffd855b2c00c90 x18: 000000000000001c [ 3.125328] x17: 000000005358439e x16: ffffd855b2c00d00 x15: 000000003be00b9b [ 3.132534] x14: 0000000088cee1a7 x13: 025733541e7bfbcb x12: 00000000000000d5 [ 3.139741] x11: 000000009d2113cd x10: 000000009d2113cd x9 : ffffd855b059298c [ 3.146947] x8 : ffffd855b2dbf980 x7 : 0000000000000000 x6 : 0000000000000002 [ 3.154154] x5 : 0000000000000000 x4 : 0000000000000002 x3 : 00000000000000d5 [ 3.161359] x2 : ffff2ab984b5c000 x1 : 0000000000000040 x0 : ffff030e00290000 [ 3.168565] Call trace: [ 3.171037] kthread_should_stop+0x2c/0x44 [ 3.175176] add_early_randomness+0xb8/0x124 [ 3.179490] hwrng_register+0x174/0x220 [ 3.183366] tpm_chip_register+0xc4/0x290 [ 3.187417] tpm_cr50_i2c_probe+0x27c/0x2c0 [ 3.191642] i2c_device_probe+0x10c/0x360 [ 3.195695] really_probe+0xc8/0x3e0 [ 3.199307] __driver_probe_device+0x84/0x190 [ 3.203706] driver_probe_device+0x44/0x120 [ 3.207929] __device_attach_driver+0xc4/0x160 [ 3.212415] bus_for_each_drv+0x80/0xe0 [ 3.216288] __device_attach+0xb0/0x1f0 [ 3.220162] device_initial_probe+0x1c/0x30 [ 3.224386] bus_probe_device+0xa4/0xb0 [ 3.228260] device_add+0x3d0/0x8d0 [ 3.231788] device_register+0x28/0x40 [ 3.235577] i2c_new_client_device+0x15c/0x29c [ 3.240067] of_i2c_register_device+0xc4/0xf0 [ 3.244465] of_i2c_register_devices+0x84/0x140 [ 3.249040] i2c_register_adapter+0x200/0x6dc [ 3.253442] __i2c_add_numbered_adapter+0x68/0xc0 [ 3.258194] i2c_add_adapter+0xb0/0xe0 [ 3.261982] mtk_i2c_probe+0x4e8/0x660 [ 3.265771] platform_probe+0x70/0xe0 [ 3.269471] really_probe+0xc8/0x3e0 [ 3.273082] __driver_probe_device+0x84/0x190 [ 3.277481] driver_probe_device+0x44/0x120 [ 3.281704] __driver_attach+0xb8/0x230 [ 3.285577] bus_for_each_dev+0x78/0xdc [ 3.289450] driver_attach+0x2c/0x3c [ 3.293061] bus_add_driver+0x184/0x240 [ 3.296935] driver_register+0x80/0x13c [ 3.300809] __platform_driver_register+0x30/0x3c [ 3.305559] mtk_i2c_driver_init+0x24/0x30 [ 3.309699] do_one_initcall+0x7c/0x3d0 [ 3.313574] kernel_init_freeable+0x308/0x378 [ 3.317977] kernel_init+0x2c/0x140 [ 3.321505] ret_from_fork+0x10/0x20 [ 3.325117] irq event stamp: 620756 [ 3.328639] hardirqs last enabled at (620755): [<ffffd855b0fc7cb4>] _raw_spin_unlock_irqrestore+0xa4/0xb0 [ 3.338382] hardirqs last disabled at (620756): [<ffffd855b0fb9724>] el1_dbg+0x24/0x90 [ 3.346373] softirqs last enabled at (618152): [<ffffd855afc10be4>] __do_softirq+0x414/0x588 [ 3.354973] softirqs last disabled at (618147): [<ffffd855afc171d8>] ____do_softirq+0x18/0x24 [ 3.363575] ---[ end trace 0000000000000000 ]--- [ 3.368278] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000 [ 3.377165] Mem abort info: [ 3.380006] ESR = 0x0000000096000004 [ 3.383824] EC = 0x25: DABT (current EL), IL = 32 bits [ 3.389203] SET = 0, FnV = 0 [ 3.392306] EA = 0, S1PTW = 0 [ 3.395494] FSC = 0x04: level 0 translation fault [ 3.400432] Data abort info: [ 3.403359] ISV = 0, ISS = 0x00000004 [ 3.407261] CM = 0, WnR = 0 [ 3.410275] [0000000000000000] user address but active_mm is swapper [ 3.416720] Internal error: Oops: 0000000096000004 [#1] SMP [ 3.422290] Modules linked in: [ 3.425338] CPU: 5 PID: 1 Comm: swapper/0 Tainted: G W 6.1.0-rc4-next-20221108+ #111 [ 3.434466] Hardware name: Acer Tomato (rev2) board (DT) [ 3.439772] pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 3.446728] pc : kthread_should_stop+0x38/0x44 [ 3.451170] lr : add_hwgenerator_randomness+0x6c/0x110 [ 3.456304] sp : ffff80000808b570 [ 3.459611] x29: ffff80000808b570 x28: 0000000000061a80 x27: 00000000000f4240 [ 3.466743] x26: ffff030f3efe9d78 x25: ffffd855b3930000 x24: 0000000000000000 [ 3.473875] x23: 0000000000000000 x22: ffffd855b2c00d08 x21: 0000000000000020 [ 3.481007] x20: 0000000000000000 x19: ffffd855b2c00c90 x18: 000000000000001c [ 3.488139] x17: 000000005358439e x16: ffffd855b2c00d00 x15: 000000003be00b9b [ 3.495270] x14: 0000000088cee1a7 x13: 025733541e7bfbcb x12: 00000000000000d5 [ 3.502402] x11: 000000009d2113cd x10: 000000009d2113cd x9 : ffffd855b059298c [ 3.509533] x8 : ffffd855b2dbf980 x7 : 0000000000000000 x6 : 0000000000000002 [ 3.516665] x5 : 0000000000000000 x4 : 0000000000000002 x3 : 00000000000000d5 [ 3.523796] x2 : ffff2ab984b5c000 x1 : 0000000000000040 x0 : 0000000000000000 [ 3.530927] Call trace: [ 3.533366] kthread_should_stop+0x38/0x44 [ 3.537459] add_early_randomness+0xb8/0x124 [ 3.541725] hwrng_register+0x174/0x220 [ 3.545557] tpm_chip_register+0xc4/0x290 [ 3.549562] tpm_cr50_i2c_probe+0x27c/0x2c0 [ 3.553739] i2c_device_probe+0x10c/0x360 [ 3.557746] really_probe+0xc8/0x3e0 [ 3.561316] __driver_probe_device+0x84/0x190 [ 3.565667] driver_probe_device+0x44/0x120 [ 3.569845] __device_attach_driver+0xc4/0x160 [ 3.574284] bus_for_each_drv+0x80/0xe0 [ 3.578115] __device_attach+0xb0/0x1f0 [ 3.581944] device_initial_probe+0x1c/0x30 [ 3.586122] bus_probe_device+0xa4/0xb0 [ 3.589952] device_add+0x3d0/0x8d0 [ 3.593437] device_register+0x28/0x40 [ 3.597183] i2c_new_client_device+0x15c/0x29c [ 3.601623] of_i2c_register_device+0xc4/0xf0 [ 3.605974] of_i2c_register_devices+0x84/0x140 [ 3.610499] i2c_register_adapter+0x200/0x6dc [ 3.614853] __i2c_add_numbered_adapter+0x68/0xc0 [ 3.619555] i2c_add_adapter+0xb0/0xe0 [ 3.623300] mtk_i2c_probe+0x4e8/0x660 [ 3.627044] platform_probe+0x70/0xe0 [ 3.630702] really_probe+0xc8/0x3e0 [ 3.634271] __driver_probe_device+0x84/0x190 [ 3.638622] driver_probe_device+0x44/0x120 [ 3.642800] __driver_attach+0xb8/0x230 [ 3.646631] bus_for_each_dev+0x78/0xdc [ 3.650460] driver_attach+0x2c/0x3c [ 3.654030] bus_add_driver+0x184/0x240 [ 3.657859] driver_register+0x80/0x13c [ 3.661690] __platform_driver_register+0x30/0x3c [ 3.666390] mtk_i2c_driver_init+0x24/0x30 [ 3.670483] do_one_initcall+0x7c/0x3d0 [ 3.674313] kernel_init_freeable+0x308/0x378 [ 3.678668] kernel_init+0x2c/0x140 [ 3.682153] ret_from_fork+0x10/0x20 [ 3.685725] Code: d65f03c0 d4210000 f9431c00 d50323bf (f9400000) [ 3.691813] ---[ end trace 0000000000000000 ]--- [ 3.696984] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b [ 3.704636] SMP: stopping secondary CPUs [ 3.708557] Kernel Offset: 0x5855a7c00000 from 0xffff800008000000 [ 3.714644] PHYS_OFFSET: 0xfffffcf300000000 [ 3.718820] CPU features: 0x30000,001700a4,6600720b [ 3.723691] Memory Limit: none [ 3.727265] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b ]--- Regards, Angelo
On Tue, Nov 08, 2022 at 11:53:23AM +0100, AngeloGioacchino Del Regno wrote: > Il 06/11/22 16:02, Jason A. Donenfeld ha scritto: > > Rather than calling add_device_randomness(), the add_early_randomness() > > function should use add_hwgenerator_randomness(), so that the early > > entropy can be potentially credited, which allows for the RNG to > > initialize earlier without having to wait for the kthread to come up. > > > > This requires some minor API refactoring, by adding a `sleep_after` > > parameter to add_hwgenerator_randomness(), so that we don't hit a > > blocking sleep from add_early_randomness(). > > > > Cc: Herbert Xu <herbert@gondor.apana.org.au> > > Cc: Dominik Brodowski <linux@dominikbrodowski.net> > > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > > Acked-by: Herbert Xu <herbert@gondor.apana.org.au> > > --- > > Herbert - it might be easiest for me to take this patch if you want? Or > > if this will interfere with what you have going on, you can take it. Let > > me know what you feel like. -Jason > > > > drivers/char/hw_random/core.c | 8 +++++--- > > drivers/char/random.c | 12 ++++++------ > > include/linux/random.h | 2 +- > > 3 files changed, 12 insertions(+), 10 deletions(-) > > > > Hello, > > I tried booting next-20221108 on Acer Tomato Chromebook (MediaTek MT8195) but > this commit is producing a kernel panic. Thanks for the report. I see exactly what the problem is, and I'll send a v+1 right away. Jason
Il 08/11/22 12:24, Jason A. Donenfeld ha scritto: > Rather than calling add_device_randomness(), the add_early_randomness() > function should use add_hwgenerator_randomness(), so that the early > entropy can be potentially credited, which allows for the RNG to > initialize earlier without having to wait for the kthread to come up. > > This requires some minor API refactoring, by adding a `sleep_after` > parameter to add_hwgenerator_randomness(), so that we don't hit a > blocking sleep from add_early_randomness(). > > Reviewed-by: Dominik Brodowski <linux@dominikbrodowski.net> > Acked-by: Herbert Xu <herbert@gondor.apana.org.au> > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> On MT8192 Asurada, MT8195 Tomato Chromebooks: Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Thanks for the fast fix! Regards, Angelo
On 08.11.2022 12:44, AngeloGioacchino Del Regno wrote: > Il 08/11/22 12:24, Jason A. Donenfeld ha scritto: >> Rather than calling add_device_randomness(), the add_early_randomness() >> function should use add_hwgenerator_randomness(), so that the early >> entropy can be potentially credited, which allows for the RNG to >> initialize earlier without having to wait for the kthread to come up. >> >> This requires some minor API refactoring, by adding a `sleep_after` >> parameter to add_hwgenerator_randomness(), so that we don't hit a >> blocking sleep from add_early_randomness(). >> >> Reviewed-by: Dominik Brodowski <linux@dominikbrodowski.net> >> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> >> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > > Reviewed-by: AngeloGioacchino Del Regno > <angelogioacchino.delregno@collabora.com> > > On MT8192 Asurada, MT8195 Tomato Chromebooks: > Tested-by: AngeloGioacchino Del Regno > <angelogioacchino.delregno@collabora.com> > > Thanks for the fast fix! I also confirm that this version fixed the boot issue observed on most of my test systems with Linux next-20221108. Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> Best regards
Greeting,
FYI, we noticed WARNING:at_kernel/kthread.c:#kthread_should_stop due to commit (built with gcc-11):
commit: cea83a6b31856293ceab2c6ebab7843322fe105e ("[PATCH v3] hw_random: use add_hwgenerator_randomness() for early entropy")
url: https://github.com/intel-lab-lkp/linux/commits/UPDATE-20221106-230344/Jason-A-Donenfeld/hw_random-use-add_hwgenerator_randomness-for-early-entropy/20221106-044656
base: https://git.kernel.org/cgit/linux/kernel/git/gregkh/char-misc.git 30a0b95b1335e12efef89dd78518ed3e4a71a763
patch subject: [PATCH v3] hw_random: use add_hwgenerator_randomness() for early entropy
in testcase: stress-ng
version: stress-ng-x86_64-0.11-06_20221109
with following parameters:
nr_threads: 100%
testtime: 60s
class: cpu
test: str
cpufreq_governor: performance
on test machine: 96 threads 2 sockets Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz (Cascade Lake) with 512G memory
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
If you fix the issue, kindly add following tag
| Reported-by: kernel test robot <oliver.sang@intel.com>
| Link: https://lore.kernel.org/oe-lkp/202211132159.cd70b1c0-oliver.sang@intel.com
[ 23.856915][ T1] ------------[ cut here ]------------
[ 23.862386][ T1] WARNING: CPU: 48 PID: 1 at kernel/kthread.c:75 kthread_should_stop (kernel/kthread.c:75 kernel/kthread.c:157)
[ 23.871363][ T1] Modules linked in:
[ 23.875261][ T1] CPU: 48 PID: 1 Comm: swapper/0 Not tainted 6.1.0-rc3-00001-gcea83a6b3185 #1
[ 23.884138][ T1] Hardware name: Intel Corporation S2600WFT/S2600WFT, BIOS SE5C620.86B.02.01.0008.031920191559 03/19/2019
[ 23.895447][ T1] RIP: 0010:kthread_should_stop (kernel/kthread.c:75 kernel/kthread.c:157)
[ 23.901164][ T1] Code: 00 0f 1f 44 00 00 65 48 8b 04 25 00 ad 01 00 f6 40 2e 20 74 15 48 8b 80 08 0a 00 00 48 8b 00 48 d1 e8 83 e0 01 c3 cc cc cc cc <0f> 0b eb e7 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 84 00 00 00 00
All code
========
0: 00 0f add %cl,(%rdi)
2: 1f (bad)
3: 44 00 00 add %r8b,(%rax)
6: 65 48 8b 04 25 00 ad mov %gs:0x1ad00,%rax
d: 01 00
f: f6 40 2e 20 testb $0x20,0x2e(%rax)
13: 74 15 je 0x2a
15: 48 8b 80 08 0a 00 00 mov 0xa08(%rax),%rax
1c: 48 8b 00 mov (%rax),%rax
1f: 48 d1 e8 shr %rax
22: 83 e0 01 and $0x1,%eax
25: c3 retq
26: cc int3
27: cc int3
28: cc int3
29: cc int3
2a:* 0f 0b ud2 <-- trapping instruction
2c: eb e7 jmp 0x15
2e: 66 66 2e 0f 1f 84 00 data16 nopw %cs:0x0(%rax,%rax,1)
35: 00 00 00 00
39: 0f .byte 0xf
3a: 1f (bad)
3b: 84 00 test %al,(%rax)
3d: 00 00 add %al,(%rax)
...
Code starting with the faulting instruction
===========================================
0: 0f 0b ud2
2: eb e7 jmp 0xffffffffffffffeb
4: 66 66 2e 0f 1f 84 00 data16 nopw %cs:0x0(%rax,%rax,1)
b: 00 00 00 00
f: 0f .byte 0xf
10: 1f (bad)
11: 84 00 test %al,(%rax)
13: 00 00 add %al,(%rax)
...
[ 23.920849][ T1] RSP: 0000:ffffc90000073b20 EFLAGS: 00010246
[ 23.926910][ T1] RAX: ffff888100df0000 RBX: 0000000000000000 RCX: 0000000000000000
[ 23.934914][ T1] RDX: 000000000000000e RSI: 0000000000000296 RDI: ffffffff830d57b8
[ 23.942914][ T1] RBP: 0000000000000296 R08: 00000000c7b95e77 R09: 00000000294e6f92
[ 23.950912][ T1] R10: 000000000defdd81 R11: 00000000f09cfad9 R12: ffff88c0882c1700
[ 23.958911][ T1] R13: 0000000000000020 R14: 0000000000000000 R15: 00000000001b15d1
[ 23.966910][ T1] FS: 0000000000000000(0000) GS:ffff88bf7fc00000(0000) knlGS:0000000000000000
[ 23.975872][ T1] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 23.982459][ T1] CR2: 0000000000000000 CR3: 000000807ee0a001 CR4: 00000000007706e0
[ 23.990463][ T1] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 23.998463][ T1] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 24.006470][ T1] PKRU: 55555554
[ 24.010006][ T1] Call Trace:
[ 24.013282][ T1] <TASK>
[ 24.016220][ T1] add_hwgenerator_randomness (drivers/char/random.c:885)
[ 24.021853][ T1] hwrng_register (drivers/char/hw_random/core.c:590)
[ 24.026528][ T1] tpm_chip_register (drivers/char/tpm/tpm-chip.c:254)
[ 24.031981][ T1] tpm_tis_core_init.cold (drivers/char/tpm/tpm_tis_core.c:1107)
[ 24.037355][ T1] tpm_tis_plat_probe (drivers/char/tpm/tpm_tis.c:324)
[ 24.042294][ T1] platform_probe (drivers/base/platform.c:1400)
[ 24.046797][ T1] really_probe (drivers/base/dd.c:560 drivers/base/dd.c:639)
[ 24.051207][ T1] ? pm_runtime_barrier (arch/x86/include/asm/paravirt.h:596 arch/x86/include/asm/qspinlock.h:57 include/linux/spinlock.h:203 include/linux/spinlock_api_smp.h:158 include/linux/spinlock.h:400 drivers/base/power/runtime.c:1413)
[ 24.056227][ T1] __driver_probe_device (drivers/base/dd.c:719 drivers/base/dd.c:776)
[ 24.061416][ T1] driver_probe_device (drivers/base/dd.c:808)
[ 24.066347][ T1] __driver_attach (drivers/base/dd.c:1191)
[ 24.071020][ T1] ? __device_attach_driver (drivers/base/dd.c:1135)
[ 24.076558][ T1] ? __device_attach_driver (drivers/base/dd.c:1135)
[ 24.082097][ T1] bus_for_each_dev (drivers/base/bus.c:301)
[ 24.086774][ T1] bus_add_driver (drivers/base/bus.c:618)
[ 24.091443][ T1] driver_register (drivers/base/driver.c:246)
[ 24.096116][ T1] ? tpm_init (drivers/char/tpm/tpm_tis.c:379)
[ 24.100263][ T1] init_tis (drivers/char/tpm/tpm_tis.c:386)
[ 24.104250][ T1] ? tpm_init (drivers/char/tpm/tpm_tis.c:379)
[ 24.108399][ T1] ? ktime_get (kernel/time/timekeeping.c:195 (discriminator 3) kernel/time/timekeeping.c:289 (discriminator 3) kernel/time/timekeeping.c:388 (discriminator 3) kernel/time/timekeeping.c:848 (discriminator 3))
[ 24.112637][ T1] ? tpm_init (drivers/char/tpm/tpm_tis.c:379)
[ 24.116789][ T1] do_one_initcall (init/main.c:1303)
[ 24.121461][ T1] do_initcalls (init/main.c:1375 init/main.c:1392)
[ 24.125784][ T1] kernel_init_freeable (init/main.c:1635)
[ 24.130978][ T1] ? rest_init (init/main.c:1511)
[ 24.135386][ T1] kernel_init (init/main.c:1521)
[ 24.139711][ T1] ret_from_fork (arch/x86/entry/entry_64.S:306)
[ 24.144122][ T1] </TASK>
[ 24.147143][ T1] ---[ end trace 0000000000000000 ]---
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
sudo bin/lkp install job.yaml # job file is attached in this email
bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run
sudo bin/lkp run generated-yaml-file
# if come across any failure that blocks the test,
# please remove ~/.lkp and /lkp dir to run from a clean state.
On Sun, Nov 13, 2022 at 09:48:46PM +0800, kernel test robot wrote: > > Greeting, > > FYI, we noticed WARNING:at_kernel/kthread.c:#kthread_should_stop due to commit (built with gcc-11): > > commit: cea83a6b31856293ceab2c6ebab7843322fe105e ("[PATCH v3] hw_random: use add_hwgenerator_randomness() for early entropy") > url: https://github.com/intel-lab-lkp/linux/commits/UPDATE-20221106-230344/Jason-A-Donenfeld/hw_random-use-add_hwgenerator_randomness-for-early-entropy/20221106-044656 > base: https://git.kernel.org/cgit/linux/kernel/git/gregkh/char-misc.git 30a0b95b1335e12efef89dd78518ed3e4a71a763 > patch subject: [PATCH v3] hw_random: use add_hwgenerator_randomness() for early entropy A bit late. v4 was posted a while ago now.
diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c index cc002b0c2f0c..63a0a8e4505d 100644 --- a/drivers/char/hw_random/core.c +++ b/drivers/char/hw_random/core.c @@ -69,8 +69,10 @@ static void add_early_randomness(struct hwrng *rng) mutex_lock(&reading_mutex); bytes_read = rng_get_data(rng, rng_fillbuf, 32, 0); mutex_unlock(&reading_mutex); - if (bytes_read > 0) - add_device_randomness(rng_fillbuf, bytes_read); + if (bytes_read > 0) { + size_t entropy = bytes_read * 8 * rng->quality / 1024; + add_hwgenerator_randomness(rng_fillbuf, bytes_read, entropy, false); + } } static inline void cleanup_rng(struct kref *kref) @@ -528,7 +530,7 @@ static int hwrng_fillfn(void *unused) /* Outside lock, sure, but y'know: randomness. */ add_hwgenerator_randomness((void *)rng_fillbuf, rc, - entropy >> 10); + entropy >> 10, true); } hwrng_fill = NULL; return 0; diff --git a/drivers/char/random.c b/drivers/char/random.c index 4591d55cb135..70ecfcfdb1d7 100644 --- a/drivers/char/random.c +++ b/drivers/char/random.c @@ -711,7 +711,7 @@ static void __cold _credit_init_bits(size_t bits) * the above entropy accumulation routines: * * void add_device_randomness(const void *buf, size_t len); - * void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy); + * void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy, bool sleep_after); * void add_bootloader_randomness(const void *buf, size_t len); * void add_vmfork_randomness(const void *unique_vm_id, size_t len); * void add_interrupt_randomness(int irq); @@ -891,11 +891,11 @@ void add_device_randomness(const void *buf, size_t len) EXPORT_SYMBOL(add_device_randomness); /* - * Interface for in-kernel drivers of true hardware RNGs. - * Those devices may produce endless random bits and will be throttled - * when our pool is full. + * Interface for in-kernel drivers of true hardware RNGs. Those devices + * may produce endless random bits, so this function will sleep for + * some amount of time after, if the sleep_after parameter is true. */ -void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy) +void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy, bool sleep_after) { mix_pool_bytes(buf, len); credit_init_bits(entropy); @@ -904,7 +904,7 @@ void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy) * Throttle writing to once every reseed interval, unless we're not yet * initialized or no entropy is credited. */ - if (!kthread_should_stop() && (crng_ready() || !entropy)) + if (!kthread_should_stop() && (crng_ready() || !entropy) && sleep_after) schedule_timeout_interruptible(crng_reseed_interval()); } EXPORT_SYMBOL_GPL(add_hwgenerator_randomness); diff --git a/include/linux/random.h b/include/linux/random.h index 2bdd3add3400..728b29ade208 100644 --- a/include/linux/random.h +++ b/include/linux/random.h @@ -17,7 +17,7 @@ void __init add_bootloader_randomness(const void *buf, size_t len); void add_input_randomness(unsigned int type, unsigned int code, unsigned int value) __latent_entropy; void add_interrupt_randomness(int irq) __latent_entropy; -void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy); +void add_hwgenerator_randomness(const void *buf, size_t len, size_t entropy, bool sleep_after); #if defined(LATENT_ENTROPY_PLUGIN) && !defined(__CHECKER__) static inline void add_latent_entropy(void)
Rather than calling add_device_randomness(), the add_early_randomness() function should use add_hwgenerator_randomness(), so that the early entropy can be potentially credited, which allows for the RNG to initialize earlier without having to wait for the kthread to come up. This requires some minor API refactoring, by adding a `sleep_after` parameter to add_hwgenerator_randomness(), so that we don't hit a blocking sleep from add_early_randomness(). Cc: Herbert Xu <herbert@gondor.apana.org.au> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> --- Herbert - it might be easiest for me to take this patch if you want? Or if this will interfere with what you have going on, you can take it. Let me know what you feel like. -Jason drivers/char/hw_random/core.c | 8 +++++--- drivers/char/random.c | 12 ++++++------ include/linux/random.h | 2 +- 3 files changed, 12 insertions(+), 10 deletions(-)