Return-Path: Date: Mon, 6 Apr 2020 14:06:01 +0200 (CEST) From: =?utf-8?Q?Cl=C3=A9ment?= Leger Message-ID: <634144036.14036712.1586174761552.JavaMail.zimbra@kalray.eu> In-Reply-To: <730c75c9-15e2-19c5-d97a-190bf1e6ffaa@st.com> References: <1585699438-14394-1-git-send-email-rishabhb@codeaurora.org> <5b1c8287-0077-87e7-9364-b1f5a104c9e3@st.com> <6261646b2e0c4d9c8a30900b2f475890@codeaurora.org> <730c75c9-15e2-19c5-d97a-190bf1e6ffaa@st.com> Subject: Re: [PATCH v2 1/2] remoteproc: Add character device interface MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable To: Arnaud Pouliquen Cc: Rishabh Bhatnagar , linux-kernel , linux-remoteproc , Bjorn Andersson , psodagud , tsoni , sidgup List-ID: Hi Arnaud, ----- On 6 Apr, 2020, at 11:01, Arnaud Pouliquen arnaud.pouliquen@st.com wr= ote: > On 4/3/20 9:13 PM, rishabhb@codeaurora.org wrote: >> On 2020-04-02 10:28, Arnaud POULIQUEN wrote: >>> Hi >>> >>> On 4/1/20 2:03 AM, Rishabh Bhatnagar wrote: >>>> Add the character device interface for userspace applications. >>>> This interface can be used in order to boot up and shutdown >>>> remote subsystems. Currently there is only a sysfs interface >>>> which the userspace clients can use. If a usersapce application >>>> crashes after booting the remote processor does not get any >>>> indication about the crash. It might still assume that the >>>> application is running. For example modem uses remotefs service >>>> to fetch data from disk/flash memory. If the remotefs service >>>> crashes, modem keeps on requesting data which might lead to a >>>> crash. Adding a character device interface makes the remote >>>> processor tightly coupled with the user space application. >>>> A crash of the application leads to a close on the file descriptors >>>> therefore shutting down the remoteproc. >>> >>> Sorry I'm late in the discussion, I hope I've gone through the whole >>> discussion so I don't reopen a closed point... >>> >>> Something here is not crystal clear to me so I'd rather share it... >>> >>> I suppose that you the automatic restart of the application is not poss= ible to >>> stop and restart the remote processor... >> Yes correct, while we wait for the application to restart we might obser= ve a >> fatal crash. >>> >>> Why this use case can not be solved by a process monitor or a service >>> in userland that detects the application crash and stop the remote >>> firmware using >>> the sysfs interface? >>> >> What happens in the case where the process monitor itself crashes? This = is >> actually the approach we follow in our downstream code. We have a centra= l entity >> in userspace that controls bootup/shutdown of some remote processors bas= ed on >> the >> votes from userspace clients. We have observed cases where this entity >> itself crashes and remote processors are left hanging. >=20 > Your description makes me feel like this patch is only a workaround of so= mething > that > should be fixed in the userland, even if i understand that hanging is one= of the > most > critical problem and have to be fixed. > For instance, how to handle several applications that interact with the r= emote > processor > ( e.g. rpmsg service applications) how to stop and restart everything. Us= ing the > char > device would probaly resolve only a part of the issue... >=20 > I'm not aware about your environment and i'm not a userland expert. But w= hat i > still not > understand why a parent process can not do the job... > I just test a simple script on my side that treat the kill -9 of an appli= cation > ("cat" in my case). This is not entirely true, if the parent process is killed with a SIGKILL, = then the process will not be able to handle anything and the remoteproc will sti= ll be running. What I understood from Rishabh patch is a way to allow a single process han= dling the rproc state. We have the same kind of need and currently, if the user application crashes, then the rproc is still running (which happens). >=20 > #start the remote firmware > cp $1 /lib/firmware/ > echo $1> /sys/class/remoteproc/remoteproc0/firmware > echo start >/sys/class/remoteproc/remoteproc0/state > #your binary > cat /dev/kmsg > # stop the remote firmware in case of crash (and potentially some other a= pps) > echo stop >/sys/class/remoteproc/remoteproc0/state >=20 This is not really "production proof" and what happens if the application i= s responsible of setting the firmware which might be jitted ?=20 And if the script receives the SIGKILL, then we are back to the same proble= m. I really think, this is a step forward an easier and reliable use of the re= moteproc on userland to guarantee a coherent rproc state even if host application crashes. Regards, Cl=C3=A9ment > Anyway, it's just my feeling, let other people give their feedback. >=20 >>> I just want to be sure that there is no alternative to this, because >>> having two ways >>> for application to shutdown the firmware seems to me confusing... >> Does making this interface optional/configurable helps? >>> >>> What about the opposite service, mean inform the application that the r= emote >>> processor is crashed? >>> Do you identify such need? or the "auto" crash recovery is sufficient? >> Auto recovery works perfectly for us. Although there is a mechanism in >> place using QMI(Qualcomm MSM interface) that can notify clients about re= mote >> processor crash. >=20 > Thanks for the information. >=20 > Regards > Arnaud >=20 >>> >>> Thanks, >>> Arnaud >>>> >>>> Signed-off-by: Rishabh Bhatnagar >>>> --- >>>> =C2=A0drivers/remoteproc/Kconfig=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 |=C2=A0=C2=A0 9 +++ >>>> =C2=A0drivers/remoteproc/Makefile=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 |=C2=A0=C2=A0 1 + >>>> =C2=A0drivers/remoteproc/remoteproc_cdev.c=C2=A0=C2=A0=C2=A0=C2=A0 | 1= 00 +++++++++++++++++++++++++++++++ >>>> =C2=A0drivers/remoteproc/remoteproc_internal.h |=C2=A0 22 +++++++ >>>> =C2=A0include/linux/remoteproc.h=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 |=C2=A0=C2=A0 2 + >>>> =C2=A05 files changed, 134 insertions(+) >>>> =C2=A0create mode 100644 drivers/remoteproc/remoteproc_cdev.c >>>> >>>> diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig >>>> index de3862c..6374b79 100644 >>>> --- a/drivers/remoteproc/Kconfig >>>> +++ b/drivers/remoteproc/Kconfig >>>> @@ -14,6 +14,15 @@ config REMOTEPROC >>>> >>>> =C2=A0if REMOTEPROC >>>> >>>> +config REMOTEPROC_CDEV >>>> +=C2=A0=C2=A0=C2=A0 bool "Remoteproc character device interface" >>>> +=C2=A0=C2=A0=C2=A0 help >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 Say y here to have a character device = interface for Remoteproc >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 framework. Userspace can boot/shutdown= remote processors through >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 this interface. >>>> + >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 It's safe to say N if you don't want t= o use this interface. >>>> + >>>> =C2=A0config IMX_REMOTEPROC >>>> =C2=A0=C2=A0=C2=A0=C2=A0 tristate "IMX6/7 remoteproc support" >>>> =C2=A0=C2=A0=C2=A0=C2=A0 depends on ARCH_MXC >>>> diff --git a/drivers/remoteproc/Makefile b/drivers/remoteproc/Makefile >>>> index e30a1b1..b7d4f77 100644 >>>> --- a/drivers/remoteproc/Makefile >>>> +++ b/drivers/remoteproc/Makefile >>>> @@ -9,6 +9,7 @@ remoteproc-y=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 +=3D remoteproc_debugfs.o >>>> =C2=A0remoteproc-y=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 +=3D remoteproc_sysfs.o >>>> =C2=A0remoteproc-y=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 +=3D remoteproc_virtio.o >>>> =C2=A0remoteproc-y=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 +=3D remoteproc_elf_loader.o >>>> +obj-$(CONFIG_REMOTEPROC_CDEV)=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0 +=3D remoteproc_cdev.o >>>> =C2=A0obj-$(CONFIG_IMX_REMOTEPROC)=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0 +=3D imx_rproc.o >>>> =C2=A0obj-$(CONFIG_MTK_SCP)=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0 +=3D mtk_scp.o mtk_scp_ipi.o >>>> =C2=A0obj-$(CONFIG_OMAP_REMOTEPROC)=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 +=3D omap_remoteproc.o >>>> diff --git a/drivers/remoteproc/remoteproc_cdev.c >>>> b/drivers/remoteproc/remoteproc_cdev.c >>>> new file mode 100644 >>>> index 0000000..8182bd1 >>>> --- /dev/null >>>> +++ b/drivers/remoteproc/remoteproc_cdev.c >>>> @@ -0,0 +1,100 @@ >>>> +// SPDX-License-Identifier: GPL-2.0-only >>>> +/* >>>> + * Character device interface driver for Remoteproc framework. >>>> + * >>>> + * Copyright (c) 2020, The Linux Foundation. All rights reserved. >>>> + */ >>>> + >>>> +#include >>>> +#include >>>> +#include >>>> +#include >>>> +#include >>>> + >>>> +#include "remoteproc_internal.h" >>>> + >>>> +#define NUM_RPROC_DEVICES=C2=A0=C2=A0=C2=A0 64 >>>> +static dev_t rproc_cdev; >>>> +static DEFINE_IDA(cdev_minor_ida); >>>> + >>>> +static int rproc_cdev_open(struct inode *inode, struct file *file) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 struct rproc *rproc; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 rproc =3D container_of(inode->i_cdev, struct rproc= , char_dev); >>>> + >>>> +=C2=A0=C2=A0=C2=A0 if (!rproc) >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return -EINVAL; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 if (rproc->state =3D=3D RPROC_RUNNING) >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return -EBUSY; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 return rproc_boot(rproc); >>>> +} >>>> + >>>> +static int rproc_cdev_release(struct inode *inode, struct file *file) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 struct rproc *rproc; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 rproc =3D container_of(inode->i_cdev, struct rproc= , char_dev); >>>> + >>>> +=C2=A0=C2=A0=C2=A0 if (!rproc || rproc->state !=3D RPROC_RUNNING) >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return -EINVAL; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 rproc_shutdown(rproc); >>>> + >>>> +=C2=A0=C2=A0=C2=A0 return 0; >>>> +} >>>> + >>>> +static const struct file_operations rproc_fops =3D { >>>> +=C2=A0=C2=A0=C2=A0 .open =3D rproc_cdev_open, >>>> +=C2=A0=C2=A0=C2=A0 .release =3D rproc_cdev_release, >>>> +}; >>>> + >>>> +int rproc_char_device_add(struct rproc *rproc) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 int ret, minor; >>>> +=C2=A0=C2=A0=C2=A0 dev_t cdevt; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 minor =3D ida_simple_get(&cdev_minor_ida, 0, NUM_R= PROC_DEVICES, >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 GFP_KERNEL); >>>> +=C2=A0=C2=A0=C2=A0 if (minor < 0) { >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 dev_err(&rproc->dev, "%s: = No more minor numbers left! rc:%d\n", >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __= func__, minor); >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return -ENODEV; >>>> +=C2=A0=C2=A0=C2=A0 } >>>> + >>>> +=C2=A0=C2=A0=C2=A0 cdev_init(&rproc->char_dev, &rproc_fops); >>>> +=C2=A0=C2=A0=C2=A0 rproc->char_dev.owner =3D THIS_MODULE; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 cdevt =3D MKDEV(MAJOR(rproc_cdev), minor); >>>> +=C2=A0=C2=A0=C2=A0 ret =3D cdev_add(&rproc->char_dev, cdevt, 1); >>>> +=C2=A0=C2=A0=C2=A0 if (ret < 0) >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 ida_simple_remove(&cdev_mi= nor_ida, minor); >>>> + >>>> +=C2=A0=C2=A0=C2=A0 rproc->dev.devt =3D cdevt; >>>> +=C2=A0=C2=A0=C2=A0 return ret; >>>> +} >>>> + >>>> +void rproc_char_device_remove(struct rproc *rproc) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 __unregister_chrdev(MAJOR(rproc->dev.devt), MINOR(= rproc->dev.devt), 1, >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 "rproc"); >>>> +=C2=A0=C2=A0=C2=A0 ida_simple_remove(&cdev_minor_ida, MINOR(rproc->de= v.devt)); >>>> +} >>>> + >>>> +void __init rproc_init_cdev(void) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 int ret; >>>> + >>>> +=C2=A0=C2=A0=C2=A0 ret =3D alloc_chrdev_region(&rproc_cdev, 0, NUM_RP= ROC_DEVICES, "rproc"); >>>> +=C2=A0=C2=A0=C2=A0 if (ret < 0) { >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 pr_err("Failed to alloc rp= roc_cdev region, err %d\n", ret); >>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return; >>>> +=C2=A0=C2=A0=C2=A0 } >>>> +} >>>> + >>>> +void __exit rproc_exit_cdev(void) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 __unregister_chrdev(MAJOR(rproc_cdev), 0, NUM_RPRO= C_DEVICES, "rproc"); >>>> +} >>>> diff --git a/drivers/remoteproc/remoteproc_internal.h >>>> b/drivers/remoteproc/remoteproc_internal.h >>>> index 493ef92..28d61a1 100644 >>>> --- a/drivers/remoteproc/remoteproc_internal.h >>>> +++ b/drivers/remoteproc/remoteproc_internal.h >>>> @@ -47,6 +47,27 @@ struct dentry *rproc_create_trace_file(const char *= name, >>>> struct rproc *rproc, >>>> =C2=A0int rproc_init_sysfs(void); >>>> =C2=A0void rproc_exit_sysfs(void); >>>> >>>> +#ifdef CONFIG_REMOTEPROC_CDEV >>>> +void rproc_init_cdev(void); >>>> +void rproc_exit_cdev(void); >>>> +int rproc_char_device_add(struct rproc *rproc); >>>> +void rproc_char_device_remove(struct rproc *rproc); >>>> +#else >>>> +static inline void rproc_init_cdev(void) >>>> +{ >>>> +} >>>> +static inline void rproc_exit_cdev(void) >>>> +{ >>>> +} >>>> +static inline int rproc_char_device_add(struct rproc *rproc) >>>> +{ >>>> +=C2=A0=C2=A0=C2=A0 return 0; >>>> +} >>>> +static inline void=C2=A0 rproc_char_device_remove(struct rproc *rproc= ) >>>> +{ >>>> +} >>>> +#endif >>>> + >>>> =C2=A0void rproc_free_vring(struct rproc_vring *rvring); >>>> =C2=A0int rproc_alloc_vring(struct rproc_vdev *rvdev, int i); >>>> >>>> @@ -63,6 +84,7 @@ struct resource_table *rproc_elf_find_loaded_rsc_tab= le(struct >>>> rproc *rproc, >>>> =C2=A0struct rproc_mem_entry * >>>> =C2=A0rproc_find_carveout_by_name(struct rproc *rproc, const char *nam= e, ...); >>>> >>>> + >>>> =C2=A0static inline >>>> =C2=A0int rproc_fw_sanity_check(struct rproc *rproc, const struct firm= ware *fw) >>>> =C2=A0{ >>>> diff --git a/include/linux/remoteproc.h b/include/linux/remoteproc.h >>>> index 16ad666..c4ca796 100644 >>>> --- a/include/linux/remoteproc.h >>>> +++ b/include/linux/remoteproc.h >>>> @@ -37,6 +37,7 @@ >>>> >>>> =C2=A0#include >>>> =C2=A0#include >>>> +#include >>>> =C2=A0#include >>>> =C2=A0#include >>>> =C2=A0#include >>>> @@ -514,6 +515,7 @@ struct rproc { >>>> =C2=A0=C2=A0=C2=A0=C2=A0 bool auto_boot; >>>> =C2=A0=C2=A0=C2=A0=C2=A0 struct list_head dump_segments; >>>> =C2=A0=C2=A0=C2=A0=C2=A0 int nb_vdev; >>>> +=C2=A0=C2=A0=C2=A0 struct cdev char_dev; >>>> =C2=A0}; >>>> >>>> =C2=A0/**