The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
Introduce a shutdown state in the optee device object to return an immediate error to all RPC requests in the shutdown path.
Fixes: f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") Signed-off-by: Lars Persson larper@axis.com --- drivers/tee/optee/optee_private.h | 1 + drivers/tee/optee/smc_abi.c | 5 ++++- drivers/tee/optee/supp.c | 8 ++++++++ 3 files changed, 13 insertions(+), 1 deletion(-)
diff --git a/drivers/tee/optee/optee_private.h b/drivers/tee/optee/optee_private.h index 46f74ab07c7e..9eb72931e11f 100644 --- a/drivers/tee/optee/optee_private.h +++ b/drivers/tee/optee/optee_private.h @@ -162,6 +162,7 @@ struct optee { struct tee_shm_pool *pool; unsigned int rpc_arg_count; bool scan_bus_done; + bool shutting_down; struct workqueue_struct *scan_bus_wq; struct work_struct scan_bus_work; }; diff --git a/drivers/tee/optee/smc_abi.c b/drivers/tee/optee/smc_abi.c index 449d6a72d289..10af747da816 100644 --- a/drivers/tee/optee/smc_abi.c +++ b/drivers/tee/optee/smc_abi.c @@ -1356,7 +1356,10 @@ static int optee_smc_remove(struct platform_device *pdev) */ static void optee_shutdown(struct platform_device *pdev) { - optee_disable_shm_cache(platform_get_drvdata(pdev)); + struct optee *optee = platform_get_drvdata(pdev); + + optee->shutting_down = true; + optee_disable_shm_cache(optee); }
static int optee_probe(struct platform_device *pdev) diff --git a/drivers/tee/optee/supp.c b/drivers/tee/optee/supp.c index 322a543b8c27..801b4ec659e8 100644 --- a/drivers/tee/optee/supp.c +++ b/drivers/tee/optee/supp.c @@ -83,6 +83,14 @@ u32 optee_supp_thrd_req(struct tee_context *ctx, u32 func, size_t num_params, bool interruptable; u32 ret;
+ /* + * When the system is shutting down we cannot talk + * to the supplicant anymore even if we seem to have + * an open session to it. + */ + if (optee->shutting_down) + return TEEC_ERROR_COMMUNICATION; + /* * Return in case there is no supplicant available and * non-blocking request.
On 2022-01-20 11:40:00, Lars Persson wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
Introduce a shutdown state in the optee device object to return an immediate error to all RPC requests in the shutdown path.
Fixes: f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") Signed-off-by: Lars Persson larper@axis.com
I think that this change looks alright to fix the deadlock but I don't feel qualified to definitively say that this is the right fix. Something makes me wonder why the session is kept alive and if there's an underlying bug that we're masking over but I don't understand the lifetime of the objects, sessions, etc., when the supplicant is involved. I'm going to defer to Jens and Sumit on the design aspects.
Tyler
drivers/tee/optee/optee_private.h | 1 + drivers/tee/optee/smc_abi.c | 5 ++++- drivers/tee/optee/supp.c | 8 ++++++++ 3 files changed, 13 insertions(+), 1 deletion(-)
diff --git a/drivers/tee/optee/optee_private.h b/drivers/tee/optee/optee_private.h index 46f74ab07c7e..9eb72931e11f 100644 --- a/drivers/tee/optee/optee_private.h +++ b/drivers/tee/optee/optee_private.h @@ -162,6 +162,7 @@ struct optee { struct tee_shm_pool *pool; unsigned int rpc_arg_count; bool scan_bus_done;
- bool shutting_down; struct workqueue_struct *scan_bus_wq; struct work_struct scan_bus_work;
}; diff --git a/drivers/tee/optee/smc_abi.c b/drivers/tee/optee/smc_abi.c index 449d6a72d289..10af747da816 100644 --- a/drivers/tee/optee/smc_abi.c +++ b/drivers/tee/optee/smc_abi.c @@ -1356,7 +1356,10 @@ static int optee_smc_remove(struct platform_device *pdev) */ static void optee_shutdown(struct platform_device *pdev) {
- optee_disable_shm_cache(platform_get_drvdata(pdev));
- struct optee *optee = platform_get_drvdata(pdev);
- optee->shutting_down = true;
- optee_disable_shm_cache(optee);
} static int optee_probe(struct platform_device *pdev) diff --git a/drivers/tee/optee/supp.c b/drivers/tee/optee/supp.c index 322a543b8c27..801b4ec659e8 100644 --- a/drivers/tee/optee/supp.c +++ b/drivers/tee/optee/supp.c @@ -83,6 +83,14 @@ u32 optee_supp_thrd_req(struct tee_context *ctx, u32 func, size_t num_params, bool interruptable; u32 ret;
- /*
* When the system is shutting down we cannot talk* to the supplicant anymore even if we seem to have* an open session to it.*/- if (optee->shutting_down)
return TEEC_ERROR_COMMUNICATION;- /*
- Return in case there is no supplicant available and
- non-blocking request.
-- 2.30.2
Hi Lars,
On Thu, 20 Jan 2022 at 16:10, Lars Persson larper@axis.com wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
This looks like another side effect (earlier known one is here [1]) of shared memory cache being allocated via RPC in a particular client's context. There is an appropriate fix from Jens here [2] to rather use driver internal tee_contex for RPC allocations. Can you try that and let us know if you still observe this issue?
[1] https://github.com/OP-TEE/optee_os/issues/1918 [2] https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro....
-Sumit
Introduce a shutdown state in the optee device object to return an immediate error to all RPC requests in the shutdown path.
Fixes: f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") Signed-off-by: Lars Persson larper@axis.com
drivers/tee/optee/optee_private.h | 1 + drivers/tee/optee/smc_abi.c | 5 ++++- drivers/tee/optee/supp.c | 8 ++++++++ 3 files changed, 13 insertions(+), 1 deletion(-)
diff --git a/drivers/tee/optee/optee_private.h b/drivers/tee/optee/optee_private.h index 46f74ab07c7e..9eb72931e11f 100644 --- a/drivers/tee/optee/optee_private.h +++ b/drivers/tee/optee/optee_private.h @@ -162,6 +162,7 @@ struct optee { struct tee_shm_pool *pool; unsigned int rpc_arg_count; bool scan_bus_done;
bool shutting_down; struct workqueue_struct *scan_bus_wq; struct work_struct scan_bus_work;}; diff --git a/drivers/tee/optee/smc_abi.c b/drivers/tee/optee/smc_abi.c index 449d6a72d289..10af747da816 100644 --- a/drivers/tee/optee/smc_abi.c +++ b/drivers/tee/optee/smc_abi.c @@ -1356,7 +1356,10 @@ static int optee_smc_remove(struct platform_device *pdev) */ static void optee_shutdown(struct platform_device *pdev) {
optee_disable_shm_cache(platform_get_drvdata(pdev));
struct optee *optee = platform_get_drvdata(pdev);optee->shutting_down = true;optee_disable_shm_cache(optee);}
static int optee_probe(struct platform_device *pdev) diff --git a/drivers/tee/optee/supp.c b/drivers/tee/optee/supp.c index 322a543b8c27..801b4ec659e8 100644 --- a/drivers/tee/optee/supp.c +++ b/drivers/tee/optee/supp.c @@ -83,6 +83,14 @@ u32 optee_supp_thrd_req(struct tee_context *ctx, u32 func, size_t num_params, bool interruptable; u32 ret;
/** When the system is shutting down we cannot talk* to the supplicant anymore even if we seem to have* an open session to it.*/if (optee->shutting_down)return TEEC_ERROR_COMMUNICATION;/* * Return in case there is no supplicant available and * non-blocking request.-- 2.30.2
On 2022-01-24 06:56, Sumit Garg wrote:
Hi Lars,
On Thu, 20 Jan 2022 at 16:10, Lars Persson larper@axis.com wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
This looks like another side effect (earlier known one is here [1]) of shared memory cache being allocated via RPC in a particular client's context. There is an appropriate fix from Jens here [2] to rather use driver internal tee_contex for RPC allocations. Can you try that and let us know if you still observe this issue?
[1] https://github.com/OP-TEE/optee_os/issues/1918 https://github.com/OP-TEE/optee_os/issues/1918 [2] https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.... https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/
-Sumit
Indeed it seems this is the root cause. Backporting the patches to 5.15 also fixes our shutdown deadlock.
Nevertheless I prefer to have this extra guard in the supplicant interface to prevent other shutdown deadlocks in the future.
/Lars
On Mon, Jan 24, 2022 at 6:48 PM Lars Persson larper@axis.com wrote:
On 2022-01-24 06:56, Sumit Garg wrote:
Hi Lars,
On Thu, 20 Jan 2022 at 16:10, Lars Persson larper@axis.com wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
This looks like another side effect (earlier known one is here [1]) of shared memory cache being allocated via RPC in a particular client's context. There is an appropriate fix from Jens here [2] to rather use driver internal tee_contex for RPC allocations. Can you try that and let us know if you still observe this issue?
[1] https://github.com/OP-TEE/optee_os/issues/1918 https://github.com/OP-TEE/optee_os/issues/1918 [2] https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.... https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/
-Sumit
Indeed it seems this is the root cause. Backporting the patches to 5.15 also fixes our shutdown deadlock.
That's good to hear.
Nevertheless I prefer to have this extra guard in the supplicant interface to prevent other shutdown deadlocks in the future.
Yes, I agree, this still makes sense. Perhaps with an updated description to avoid unnecessary confusion when applied on top of my "tee: shared memory updates" patches.
Thanks, Jens
On Tue, 25 Jan 2022 at 01:57, Jens Wiklander jens.wiklander@linaro.org wrote:
On Mon, Jan 24, 2022 at 6:48 PM Lars Persson larper@axis.com wrote:
On 2022-01-24 06:56, Sumit Garg wrote:
Hi Lars,
On Thu, 20 Jan 2022 at 16:10, Lars Persson larper@axis.com wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
This looks like another side effect (earlier known one is here [1]) of shared memory cache being allocated via RPC in a particular client's context. There is an appropriate fix from Jens here [2] to rather use driver internal tee_contex for RPC allocations. Can you try that and let us know if you still observe this issue?
[1] https://github.com/OP-TEE/optee_os/issues/1918 https://github.com/OP-TEE/optee_os/issues/1918 [2] https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.... https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/
-Sumit
Indeed it seems this is the root cause. Backporting the patches to 5.15 also fixes our shutdown deadlock.
That's good to hear.
Nevertheless I prefer to have this extra guard in the supplicant interface to prevent other shutdown deadlocks in the future.
Yes, I agree, this still makes sense. Perhaps with an updated description to avoid unnecessary confusion when applied on top of my "tee: shared memory updates" patches.
That's fine with me and we should drop the fixes tag as well.
-Sumit
Thanks, Jens
On 2022-01-25 07:05, Sumit Garg wrote:
On Tue, 25 Jan 2022 at 01:57, Jens Wiklander jens.wiklander@linaro.org wrote:
On Mon, Jan 24, 2022 at 6:48 PM Lars Persson larper@axis.com wrote:
On 2022-01-24 06:56, Sumit Garg wrote:
Hi Lars,
On Thu, 20 Jan 2022 at 16:10, Lars Persson larper@axis.com wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
This looks like another side effect (earlier known one is here [1]) of shared memory cache being allocated via RPC in a particular client's context. There is an appropriate fix from Jens here [2] to rather use driver internal tee_contex for RPC allocations. Can you try that and let us know if you still observe this issue?
https://github.com/OP-TEE/optee_os/issues/1918
https://github.com/OP-TEE/optee_os/issues/1918>
[2] https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro....
https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/
<https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro....
https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/>
-Sumit
Indeed it seems this is the root cause. Backporting the patches to 5.15 also fixes our shutdown deadlock.
That's good to hear.
Nevertheless I prefer to have this extra guard in the supplicant interface to prevent other shutdown deadlocks in the future.
Yes, I agree, this still makes sense. Perhaps with an updated description to avoid unnecessary confusion when applied on top of my "tee: shared memory updates" patches.
That's fine with me and we should drop the fixes tag as well.
Should we really drop the Fixes ? Depending on if and when the SHM owning context patch is backported, this fix may still be relevant for the 5.15 LTS branch.
/Lars
On Wed, 26 Jan 2022 at 16:18, Lars Persson larper@axis.com wrote:
On 2022-01-25 07:05, Sumit Garg wrote:
On Tue, 25 Jan 2022 at 01:57, Jens Wiklander jens.wiklander@linaro.org wrote:
On Mon, Jan 24, 2022 at 6:48 PM Lars Persson larper@axis.com wrote:
On 2022-01-24 06:56, Sumit Garg wrote:
Hi Lars,
On Thu, 20 Jan 2022 at 16:10, Lars Persson larper@axis.com wrote:
The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot") introduced a kernel shutdown regression that can be triggered after running the xtest suites.
Once the shutdown hook is called it is not possible to communicate any more with the supplicant process because the system is not scheduling task any longer. Thus if the optee driver shutdown path receives a supplicant RPC request from the OP-TEE we will deadlock the kernel's shutdown.
This unexpected event will in fact occur after the xtest suite has been run. It seems some cached SHM kept alive a context object which in turn kept alive a session towards a PTA or TA. Closing the session results in a socket RPC command being sent back from OP-TEE.
This sequence of events is captured by a 5.15 kernel annotated with extra prints:
Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001079380 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8001CC5580 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308A80 Calling OPTEE_SMC_DISABLE_SHM_CACHE OPTEE_SMC_DISABLE_SHM_CACHE returned 0 freeing SHM ptr 0xFFFFFF8006308B00 optee: optee_handle_rpc: a0=0XFFFF0000 a1=0XA0 a2=0X0 optee: optee_handle_rpc: a0=0XFFFF0005 a1=0XFFFFFF80 a2=0X61E6500 optee: handle_rpc_func_cmd: cmd = 0XA optee_supp_thrd_req: func=0XA
This looks like another side effect (earlier known one is here [1]) of shared memory cache being allocated via RPC in a particular client's context. There is an appropriate fix from Jens here [2] to rather use driver internal tee_contex for RPC allocations. Can you try that and let us know if you still observe this issue?
https://github.com/OP-TEE/optee_os/issues/1918
https://github.com/OP-TEE/optee_os/issues/1918>
[2] https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro....
https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/
<https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro....
https://lore.kernel.org/lkml/20220114150824.3578829-8-jens.wiklander@linaro.org/>
-Sumit
Indeed it seems this is the root cause. Backporting the patches to 5.15 also fixes our shutdown deadlock.
That's good to hear.
Nevertheless I prefer to have this extra guard in the supplicant interface to prevent other shutdown deadlocks in the future.
Yes, I agree, this still makes sense. Perhaps with an updated description to avoid unnecessary confusion when applied on top of my "tee: shared memory updates" patches.
That's fine with me and we should drop the fixes tag as well.
Should we really drop the Fixes ? Depending on if and when the SHM owning context patch is backported, this fix may still be relevant for the 5.15 LTS branch.
I rather consider this as a robustness patch where the actual problem is being fixed by Jens patch. I hope we should be able to target Jens patch for 5.17 as a fix which should be backported to affected stable releases.
-Sumit
/Lars
op-tee@lists.trustedfirmware.org
-
Jens Wiklander -
Lars Persson -
Sumit Garg -
Tyler Hicks