kexec on rk3399
Subsystems (3)
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.
It turns out that kexec has never worked reliably on this platform, i was just lucky until recently.
Please, can you provide some directions on how to debug the issue?
The latest test performed is:
1.- Boot v5.3-rc1
2.- Kexec into v5.2.2
3.- The kernel reports nothing (earlycon not enabled) and does not boot. The same kernel v5.2.2 works fine when booted without kexec.
Regards,
Vicenç.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Quoted 4 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.
I thought that you resolved this issue by upgrading u-boot. Did that not actually work?
Will
Quoted 10 lines · Will Deacon
Quoted 4 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.I thought that you resolved this issue by upgrading u-boot. Did that not actually work?
Will
I am now using u-boot, but that is not a solution.
My intention is to use it on the Sapphire board and on gru-kevin.
I would like the bootloader to support HDMI on the sapphire and EDP on kevin for the interactive boot menu.
U-Boot on the Sapphire boots, but has no graphic display. Kevin is not supported at all. I've achived some progress on the u-boot front, but not enough to have display.
With a kexec-based bootloader display is already working, out of the box.
From another point of view, I would like to reduce the TCB (trusted computing base). Linux is huge, but is already in the TCB, so, executing it twice does not add to the TCB. U-Boot, instead, is enlarging the TCB, in this case.
Regards,
Vicenç.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Quoted 8 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, i was just lucky until recently.
Quoted 7 lines
Please, can you provide some directions on how to debug the issue?
The latest test performed is:
1.- Boot v5.3-rc1
2.- Kexec into v5.2.2
3.- The kernel reports nothing (earlycon not enabled) and does not boot. The same kernel v5.2.2 works fine when booted without kexec.
Is there a known kernel where kexec was working (v5.2 e.g.)?
If you can find one, you can run git bisect to find the commit that breaks things.
Apart from that you should try to enable earlycon on your kernel you want to kexec into, so at least you can see if it crashes in the new kernel. If you still don't see anything you can try to use prugatory printing (if you use kexec_load and not kexec_file_load). I have a work-in-progress patches that I can share with you if needed. Just let me know.
Regards,
Matthias
Quoted 9 lines
Regards,
Vicenç._______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
_______________________________________________
Linux-rockchip mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-rockchip
Quoted 11 lines · Matthias Brugger
Quoted 6 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, ...
Is there a known kernel where kexec was working (v5.2 e.g.)?
If you can find one, you can run git bisect to find the commit that breaks things.
Yes, most kernels i have tried before 5.2 worked, more or less.
From time to time there were random freezes at boot time.
Re-booting 2 or 3 times fixed it. This was happenning about 1 out of 10 times. I've rarely seen kernel traces on dmesg, but these were not affecting the stability of the system. I've also used 5.2 (a known affected version) during days with no issues.
So, the randomness of the issue reproducibility makes bisecting too dificult.
Quoted 2 lines
Apart from that you should try to enable earlycon on your kernel you want to kexec into, so at least you can see if it crashes in the new kernel.
Added earlycon to the cmdline. Now it boots fine, although it is not working, the d_lookup issues are still there.
Basically, it looks like anything that changes the kernel memory layout (like a single extra alloc) affects the issue, making it completely random.
Quoted 4 lines
If you
still don't see anything you can try to use prugatory printing (if you usekexec_loadand notkexec_file_load). I have a work-in-progress patches that I can share with you if needed. Just let me know.
What is 'prugatory printing' ? I am using kexec_load from kexec-tools. The option to use kexec_file_load is not used.
Quoted 11 lines
Regards,
MatthiasQuoted 7 lines
Regards,
Vicenç._______________________________________________
linux-arm-kernel mailing list
[email protected] ...
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Quoted 5 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.
Quoted 7 lines
It turns out that kexec has never worked reliably on this platform, i was just lucky until recently.
Please, can you provide some directions on how to debug the issue?
The latest test performed is:
1.- Boot v5.3-rc1
Not the most trusted version, but hey, why not...
Quoted 3 lines
2.- Kexec into v5.2.2
3.- The kernel reports nothing (earlycon not enabled) and does not boot.
The same kernel v5.2.2 works fine when booted without kexec.
Can you please enable earlycon and let us know how far it goes? Your previous reports hinting at runtime memory corruption, but this seems to be much more radical...
FWIW, I was able to kexec using 5.2-rcx (for values of x around 6) on a NanoPC-T4, using a mainline u-boot and firmware.
Thanks,
M.
Signature · 1 line
Jazz is not dead. It just smells funny...
Quoted 10 lines · Marc Zyngier
Quoted 6 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, ...
Not the most trusted version, but hey, why not...
Quoted 7 lines
Quoted 3 lines
2.- Kexec into v5.2.2
3.- The kernel reports nothing (earlycon not enabled) and does not boot.
The same kernel v5.2.2 works fine when booted without kexec.Can you please enable earlycon and let us know how far it goes? Your previous reports hinting at runtime memory corruption, but this seems to be much more radical...
Details on previous email, but basically, earlycon affects reproducibility.
If it is a runtime memory corruption, what are we hunting for? an issue in the kernel before kexec or in the kernel after kexec?
Quoted 6 lines
FWIW, I was able to kexec using 5.2-rcx (for values of x around 6) on a NanoPC-T4, using a mainline u-boot and firmware.
Thanks,
M.
Regards,
Vicenç.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Quoted 15 lines · Vicente Bergas
Quoted 14 lines · Marc Zyngier
Quoted 6 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, ...
Not the most trusted version, but hey, why not...
Quoted 3 lines
2.- Kexec into v5.2.2
3.- The kernel reports nothing (earlycon not enabled) and does not boot. The same kernel v5.2.2 works fine when booted without kexec.
Quoted 8 lines
Quoted 3 lines
Can you please enable earlycon and let us know how far it goes? Your previous reports hinting at runtime memory corruption, but this seems to be much more radical...
Details on previous email, but basically, earlycon affects reproducibility.
If it is a runtime memory corruption, what are we hunting for? an issue in the kernel before kexec or in the kernel after kexec?
The obvious culprit would be DMA devices left running by the first kernel scribbling over the second kernel's memory before it's had the chance to reset them. Since boot-time memory allocation patterns tend to be relatively repeatable for a given platform and kernel configuration, "random" may well look a lot less random than you might expect, and it wouldn't be unheard of for e.g. the second kernel to mostly allocate its dentry cache from the same area the first kernel was mostly putting a network Rx descriptor ring.
Robin.
Quoted 8 lines
Quoted 7 lines
FWIW, I was able to kexec using 5.2-rcx (for values of x around 6) on a NanoPC-T4, using a mainline u-boot and firmware.
Thanks,
M.
Quoted 8 lines
Regards,
Vicenç._______________________________________________
Linux-rockchip mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-rockchip
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Quoted 6 lines · Robin Murphy
Quoted 1 line · Vicente Bergas
On Monday, July 22, 2019 4:54:41 PM CEST, Marc Zyngier wrote: ...
The obvious culprit would be DMA devices left running by the first kernel scribbling over the second kernel's memory before it's had the chance to reset them. Since boot-time memory allocation patterns tend to be relatively repeatable for a given platform and kernel configuration, "random" may well look a lot less random than you might expect, and it wouldn't be unheard of for e.g. the second kernel to mostly allocate its dentry cache from the same area the first kernel was mostly putting a network Rx descriptor ring.
Robin.
Here is another attempted test: on the first kernel disable:
# CONFIG_ZONE_DMA32 is not set
# CONFIG_DMADEVICES is not set
# CONFIG_PL330_DMA is not set
that is, all i could disable matching CONFIG_*DMA*=y, which is not much. Still enabled are:CONFIG_HAVE_DMA_CONTIGUOUS=yCONFIG_DMA_SHARED_BUFFER=yCONFIG_SCSI_DMA=yCONFIG_IOMMU_DMA=yCONFIG_HAS_DMA=yCONFIG_NEED_SG_DMA_LENGTH=yCONFIG_NEED_DMA_MAP_STATE=yCONFIG_ARCH_DMA_ADDR_T_64BIT=yCONFIG_DMA_DECLARE_COHERENT=y CONFIG_ARCH_HAS_SETUP_DMA_OPS=y CONFIG_ARCH_HAS_TEARDOWN_DMA_OPS=y CONFIG_ARCH_HAS_SYNC_DMA_FOR_DEVICE=y CONFIG_ARCH_HAS_SYNC_DMA_FOR_CPU=y CONFIG_ARCH_HAS_DMA_PREP_COHERENT=y CONFIG_ARCH_HAS_DMA_COHERENT_TO_PFN=y CONFIG_ARCH_HAS_DMA_MMAP_PGPROT=yCONFIG_DMA_REMAP=yCONFIG_DMA_DIRECT_REMAP=y
Then the second kernel still fails with d_lookup errors.
Regards,
Vicenç.
_______________________________________________
Linux-rockchip mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-rockchip
Quoted 33 lines · Vicente Bergas
Quoted 6 lines · Robin Murphy
Quoted 1 line · Vicente Bergas
On Monday, July 22, 2019 4:54:41 PM CEST, Marc Zyngier wrote: ...
The obvious culprit would be DMA devices left running by the first kernel scribbling over the second kernel's memory before it's had the chance to reset them. Since boot-time memory allocation patterns tend to be relatively repeatable for a given platform and kernel configuration, "random" may well look a lot less random than you might expect, and it wouldn't be unheard of for e.g. the second kernel to mostly allocate its dentry cache from the same area the first kernel was mostly putting a network Rx descriptor ring.
Robin.
Here is another attempted test: on the first kernel disable:
#CONFIG_ZONE_DMA32is not set
#CONFIG_DMADEVICESis not set
#CONFIG_PL330_DMAis not set
that is, all i could disable matching CONFIG_*DMA*=y, which is not much. Still enabled are:CONFIG_HAVE_DMA_CONTIGUOUS=yCONFIG_DMA_SHARED_BUFFER=yCONFIG_SCSI_DMA=yCONFIG_IOMMU_DMA=yCONFIG_HAS_DMA=yCONFIG_NEED_SG_DMA_LENGTH=yCONFIG_NEED_DMA_MAP_STATE=yCONFIG_ARCH_DMA_ADDR_T_64BIT=yCONFIG_DMA_DECLARE_COHERENT=yCONFIG_ARCH_HAS_SETUP_DMA_OPS=yCONFIG_ARCH_HAS_TEARDOWN_DMA_OPS=yCONFIG_ARCH_HAS_SYNC_DMA_FOR_DEVICE=yCONFIG_ARCH_HAS_SYNC_DMA_FOR_CPU=yCONFIG_ARCH_HAS_DMA_PREP_COHERENT=yCONFIG_ARCH_HAS_DMA_COHERENT_TO_PFN=yCONFIG_ARCH_HAS_DMA_MMAP_PGPROT=yCONFIG_DMA_REMAP=yCONFIG_DMA_DIRECT_REMAP=y
Quoted 1 line
Then the second kernel still fails with
d_lookuperrors.
Yeah, none of those configs are particularly relevant - I'm thinking more about the drivers for ethernet, wifi, USB, PCI devices, and any other peripherals which may be set up to make DMA accesses based on external stimuli and don't get shut down cleanly by a kexec.
Robin.
Quoted 9 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, i was just lucky until recently.
Please, can you provide some directions on how to debug the issue?
Thank you all for your suggestions on where the issue could be.
It seems that it was the USB driver.
Now using v5.2.8 booted with kexec from v5.2.8 with a workaround and so far so good. It is being tested on the Sapphire board.
The workaround is:
drivers/usb/dwc3/dwc3-of-simple.c +0 -0
@@ -133,6 +133,13 @@
return 0; }
+static void dwc3_of_simple_shutdown(struct platform_device *pdev) +{
+ struct dwc3_of_simple *simple = platform_get_drvdata(pdev);
+
+ reset_control_assert(simple->resets);
+}
+
static int __maybe_unused dwc3_of_simple_runtime_suspend(struct device *dev)
{
struct dwc3_of_simple *simple = dev_get_drvdata(dev);
@@ -190,6 +197,7 @@
static struct platform_driver dwc3_of_simple_driver = {
.probe = dwc3_of_simple_probe,
.remove = dwc3_of_simple_remove,
+ .shutdown = dwc3_of_simple_shutdown,
.driver = {
.name = "dwc3-of-simple",
.of_match_table = of_dwc3_simple_match,If this patch is OK after review i can resubmit it as a pull request. Should a similar change be applied to drivers/usb/dwc3/core.c ?
Regards,
Vicenç.
Hi,
Quoted 44 lines · Vicente Bergas
Quoted 9 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, i was just lucky until recently.
Please, can you provide some directions on how to debug the issue?
Thank you all for your suggestions on where the issue could be.
It seems that it was the USB driver.
Now using v5.2.8 booted with kexec from v5.2.8 with a workaround and so far so good. It is being tested on the Sapphire board.The workaround is:
drivers/usb/dwc3/dwc3-of-simple.c +7
-0@@ -133,6 +133,13 @@return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +reset_control_assert (simple->resets); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev) {drivers/usb/dwc3/dwc3-of-simple.c +1
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);@@ -190,6 +197,7 @@static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
Quoted 1 line
If this patch is OK after review i can resubmit it as a pull request.
not a pull request, just send a patch using git send-email
Quoted 1 line
Should a similar change be applied to
drivers/usb/dwc3/core.c?
Is it necessary? We haven't had any bug reports regarding that. Also, if we have reset control support in the core driver, why do we need it in of_simple? Seems like of_simple could just rely on what core does.
Signature · 1 line
balbi
_______________________________________________
Linux-rockchip mailing list [email protected]
http://lists.infradead.org/mailman/listinfo/linux-rockchip
Quoted 48 lines · Felipe Balbi
Hi,
Quoted 41 lines · Vicente Bergas
Quoted 6 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, ...
Thank you all for your suggestions on where the issue could be.
It seems that it was the USB driver.
Now using v5.2.8 booted with kexec from v5.2.8 with a workaround and so far so good. It is being tested on the Sapphire board.The workaround is:
drivers/usb/dwc3/dwc3-of-simple.c +7
-0@@ -133,6 +133,13 @@return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +reset_control_assert (simple->resets); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev) {drivers/usb/dwc3/dwc3-of-simple.c +1
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);@@ -190,6 +197,7 @@static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,Quoted 1 line
If this patch is OK after review i can resubmit it as a pull request.
not a pull request, just send a patch using git send-email
Quoted 5 lines
Quoted 1 line
Should a similar change be applied to
drivers/usb/dwc3/core.c?Is it necessary? We haven't had any bug reports regarding that. Also, if we have reset control support in the core driver, why do we need it in
of_simple? Seems likeof_simplecould just rely on what core does.
the workaround has been tested patching only core.c with
drivers/usb/dwc3/core.c +0 -0
@@ -1561,6 +1561,13 @@
return 0; }
+static void dwc3_shutdown(struct platform_device *pdev) +{
+ struct dwc3 *dwc = platform_get_drvdata(pdev);
+
+ reset_control_assert(dwc->reset);
+}
+
#ifdef CONFIG_PM
static int dwc3_core_init_for_resume(struct dwc3 *dwc) {
@@ -1866,6 +1873,7 @@
static struct platform_driver dwc3_driver = {
.probe = dwc3_probe,
.remove = dwc3_remove,
+ .shutdown = dwc3_shutdown,
.driver = {
.name = "dwc3",
.of_match_table = of_match_ptr(of_dwc3_match),and leaving dwc3-of-simple.c as is, the issue persisted.
Regards,
Vicenç.
Hi,
Quoted 78 lines · Vicente Bergas
Quoted 46 lines
Quoted 41 lines · Vicente Bergas
Quoted 6 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, ...
Thank you all for your suggestions on where the issue could be.
It seems that it was the USB driver.
Now using v5.2.8 booted with kexec from v5.2.8 with a workaround and so far so good. It is being tested on the Sapphire board.The workaround is:
drivers/usb/dwc3/dwc3-of-simple.c +7
-0@@ -133,6 +133,13 @@return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +reset_control_assert (simple->resets); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev) {drivers/usb/dwc3/dwc3-of-simple.c +1
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);@@ -190,6 +197,7 @@static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,Quoted 1 line
If this patch is OK after review i can resubmit it as a pull request.
not a pull request, just send a patch using git send-email
Quoted 5 lines
Quoted 1 line
Should a similar change be applied to
drivers/usb/dwc3/core.c?Is it necessary? We haven't had any bug reports regarding that. Also, if we have reset control support in the core driver, why do we need it in
of_simple? Seems likeof_simplecould just rely on what core does.the workaround has been tested patching only
core.cwithdrivers/usb/dwc3/core.c +8
-0@@ -1561,6 +1561,13 @@return 0 ; } +static void dwc3_shutdown (struct platform_device *pdev) +{ +struct dwc3 *dwc =platform_get_drvdata (pdev); + +reset_control_assert (dwc->reset); +} +#ifdef CONFIG_PM static int dwc3_core_init_for_resume (struct dwc3 *dwc) {@@ -1866,6 +1873,7 @@static struct platform_driver dwc3_driver = { .probe = dwc3_probe, .remove = dwc3_remove, + .shutdown = dwc3_shutdown, .driver = { .name ="dwc3" , .of_match_table =of_match_ptr (of_dwc3_match),
Quoted 1 line
and leaving
dwc3-of-simple.cas is, the issue persisted.
That's because your reset controller is not passed to dwc3 core, only to your glue layer.
Signature · 1 line
balbi
Quoted 43 lines · Vicente Bergas
Quoted 9 lines · Vicente Bergas
Hi, i have been running linux on rk3399 booted with kexec fine until 5.2 From 5.2 onwards, there are memory corruption issues as reported here:
http://lkml.iu.edu/hypermail/linux/kernel/1906.2/07211.html
kexec has been identified as the principal reason for the issues.It turns out that kexec has never worked reliably on this platform, i was just lucky until recently.
Please, can you provide some directions on how to debug the issue?
Thank you all for your suggestions on where the issue could be.
It seems that it was the USB driver.
Now using v5.2.8 booted with kexec from v5.2.8 with a workaround and so far so good. It is being tested on the Sapphire board.The workaround is:
drivers/usb/dwc3/dwc3-of-simple.c +8
-0@@ -133,6 +133,13 @@ return 0; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ + struct dwc3_of_simple *simple = platform_get_drvdata(pdev); + + reset_control_assert(simple->resets); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev) { struct dwc3_of_simple *simple = dev_get_drvdata(dev);@@ -190,6 +197,7 @@static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name = "dwc3-of-simple", .of_match_table = of_dwc3_simple_match,
Quoted 2 lines
If this patch is OK after review i can resubmit it as a pull request. Should a similar change be applied to
drivers/usb/dwc3/core.c?
This particular change looks like it's implicitly specific to RK3399, which wouldn't be ideal. Presumably if the core dwc3 driver implemented shutdown correctly (echoing parts of dwc3_remove(), I guess) then the glue layers shouldn't need anything special anyway.
Robin.
Quoted 6 lines · Robin Murphy
Quoted 1 line · Vicente Bergas
On Monday, July 22, 2019 4:31:27 PM CEST, Vicente Bergas wrote: ...
This particular change looks like it's implicitly specific to RK3399, which wouldn't be ideal. Presumably if the core dwc3 driver implemented shutdown correctly (echoing parts of
dwc3_remove(), I guess) then the glue layers shouldn't need anything special anyway.Robin.
I just checked simple->resets from dwc3-of-simple.c and it is an array with multiple resets whereas dwc->reset from core.c is NULL. So the reset seems specific to the glue layers.
Is there another way than resetting the thing that is generic enough to go to core.c and allows kexec?
Hi,
Quoted 10 lines · Vicente Bergas
Quoted 9 lines · Robin Murphy
Quoted 1 line · Vicente Bergas
On Monday, July 22, 2019 4:31:27 PM CEST, Vicente Bergas wrote: ...
This particular change looks like it's implicitly specific to RK3399, which wouldn't be ideal. Presumably if the core dwc3 driver implemented shutdown correctly (echoing parts of
dwc3_remove(), I guess) then the glue layers shouldn't need anything special anyway.
Quoted 7 lines
Quoted 1 line
Robin.
I just checked
simple->resetsfromdwc3-of-simple.cand it is an array with multiple resets whereasdwc->resetfromcore.cis NULL. So the reset seems specific to the glue layers.
Is there another way than resetting the thing that is generic enough to go tocore.cand allows kexec?
This is a really odd 'failure'. We do full soft reset during driver initialization on dwc3. We shouldn't need to assert reset on shutdown, really.
I think the problem is here:
if (simple->pulse_resets) {
ret = reset_control_reset(simple->resets);
if (ret)
goto err_resetc_put;
} else {
ret = reset_control_deassert(simple->resets);
if (ret)
goto err_resetc_put;
}Note that if pulse_resets is set, we will run a reset. But if pulse_resets is false and need_reset is true, we deassert the reset.
I think below patch is enough:
drivers/usb/dwc3/dwc3-of-simple.c +9 -4
@@ -72,7 +72,15 @@ static int dwc3_of_simple_probe(struct platform_device *pdev) ret =reset_control_reset (simple->resets);if (ret) goto err_resetc_put; - }else { + } + +if (simple->need_reset) { + ret =reset_control_assert (simple->resets); +if (ret) + goto err_resetc_put; + +usleep_range (1000 ,2000 ); + ret =reset_control_deassert (simple->resets);if (ret) goto err_resetc_put;
@@ -121,9 +129,6 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)clk_bulk_put_all (simple->num_clocks, simple->clks); simple->num_clocks =0 ; -if (!simple->pulse_resets) -reset_control_assert (simple->resets); -reset_control_put (simple->resets);pm_runtime_disable (dev);
Can you test?
Signature · 1 line
balbi
Quoted 23 lines · Felipe Balbi
Hi,
Quoted 10 lines · Vicente Bergas
Quoted 9 lines · Robin Murphy
Quoted 1 line · Vicente Bergas
On Monday, July 22, 2019 4:31:27 PM CEST, Vicente Bergas wrote: ...
This particular change looks like it's implicitly specific to RK3399, which wouldn't be ideal. Presumably if the core dwc3 driver implemented shutdown correctly (echoing parts of
dwc3_remove(), I guess) then the glue layers shouldn't need anything special anyway.Quoted 7 lines
Quoted 1 line
Robin.
I just checked
simple->resetsfromdwc3-of-simple.cand it is an array with multiple resets whereasdwc->resetfromcore.cis NULL. So the reset seems specific to the glue layers.
Is there another way than resetting the thing that is generic enough to go tocore.cand allows kexec?
Quoted 3 lines
This is a really odd 'failure'. We do full soft reset during driver initialization on dwc3. We shouldn't need to assert reset on shutdown, really.
Probing/initialisation has never been the problem. The issue for the kexec case is that when the first kernel shuts down, there is currently nothing to quiesce the controller (since only driver->shutdown gets called, not driver->remove), and thus (presumably) external USB activity causes it to keep writing back over the memory where the descriptors/command ring used to be while the second kernel boots. The second kernel will eventually probe and reset it appropriately, but by that time the damage is already done.
Yanking on a hardware reset line when the first kernel shuts down is certainly one way to stop any memory accesses if such a control is available, but presumably there's a general software way to gracefully disable the controller's DMA functions until a subsequent probe can fully reset it again - I think that would be the preferable solution.
Robin.
Quoted 51 lines
I think the problem is here:
if (simple->pulse_resets) { ret = reset_control_reset(simple->resets); if (ret) goto err_resetc_put; } else { ret = reset_control_deassert(simple->resets); if (ret) goto err_resetc_put; }Note that if
pulse_resetsis set, we will run a reset. But ifpulse_resetsis false andneed_resetis true, we deassert the reset.I think below patch is enough:
drivers/usb/dwc3/dwc3-of-simple.c +9
-4@@ -72,7 +72,15 @@ static int dwc3_of_simple_probe(struct platform_device *pdev) ret =reset_control_reset (simple->resets);if (ret) goto err_resetc_put; - }else { + } + +if (simple->need_reset) { + ret =reset_control_assert (simple->resets); +if (ret) + goto err_resetc_put; + +usleep_range (1000 ,2000 ); + ret =reset_control_deassert (simple->resets);if (ret) goto err_resetc_put;@@ -121,9 +129,6 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)clk_bulk_put_all (simple->num_clocks, simple->clks); simple->num_clocks =0 ; -if (!simple->pulse_resets) -reset_control_assert (simple->resets); -reset_control_put (simple->resets);pm_runtime_disable (dev);Can you test?
PATCH 1/1 usb: dwc3: Add shutdown to platform_driver
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
--- drivers/usb/dwc3/dwc3-of-simple.c | 6 ++++++ 1 file changed, 6 insertions(+)
Hi Felipe, Robin,
this version calls 'remove' from 'shutdown' instead of just asserting a reset because it looks like a cleaner way to stop the device.
Calling remove from shutdown in core.c instead of dwc3-of-simple.c does not fix the issue either.
It has been tested on the sapphire board, a RK3399 platform.
Regards,
Vicenç.
drivers/usb/dwc3/dwc3-of-simple.c +6 -0
@@ -133,6 +133,11 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +dwc3_of_simple_remove (pdev); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev) {struct dwc3_of_simple *simple =dev_get_drvdata (dev);
@@ -190,6 +195,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
Signature · 7 lines
2.22.1 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Quoted 39 lines · Vicente Bergas
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
Signed-off-by Vicente Bergas <[email protected]>--- drivers/usb/dwc3/dwc3-of-simple.c | 6 ++++++ 1 file changed, 6 insertions(+)Hi Felipe, Robin,
this version calls 'remove' from 'shutdown' instead of just asserting a reset because it looks like a cleaner way to stop the device.Calling remove from shutdown in
core.cinstead ofdwc3-of-simple.cdoes not fix the issue either.It has been tested on the sapphire board, a RK3399 platform.
Regards,
Vicenç.drivers/usb/dwc3/dwc3-of-simple.c +5
-0@@ -133,6 +133,11 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +dwc3_of_simple_remove (pdev); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev){
drivers/usb/dwc3/dwc3-of-simple.c +0
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);
drivers/usb/dwc3/dwc3-of-simple.c MODULE_DEVICE_TABLE(of, of_dwc3_simple_match); · line 201
Quoted 7 lines
drivers/usb/dwc3/dwc3-of-simple.c +1
-0@@ -190,6 +195,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
Hi,
please, can you provide some feedback on this?
Regards,
Vicenç.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Hi,
Quoted 31 lines · Vicente Bergas
Quoted 30 lines · Vicente Bergas
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
Signed-off-by Vicente Bergas <[email protected]>--- drivers/usb/dwc3/dwc3-of-simple.c | 6 ++++++ 1 file changed, 6 insertions(+)Hi Felipe, Robin,
this version calls 'remove' from 'shutdown' instead of just asserting a reset because it looks like a cleaner way to stop the device.Calling remove from shutdown in
core.cinstead ofdwc3-of-simple.cdoes not fix the issue either.It has been tested on the sapphire board, a RK3399 platform.
Regards,
Vicenç.drivers/usb/dwc3/dwc3-of-simple.c +0
-0@@ -133,6 +133,11 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)return 0 ; }
Quoted 14 lines
Quoted 10 lines
drivers/usb/dwc3/dwc3-of-simple.c +5
-0+static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +dwc3_of_simple_remove (pdev); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev){
drivers/usb/dwc3/dwc3-of-simple.c +0
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);drivers/usb/dwc3/dwc3-of-simple.c MODULE_DEVICE_TABLE(of, of_dwc3_simple_match); · line 198
Quoted 4 lines
drivers/usb/dwc3/dwc3-of-simple.c +1
-0@@ -190,6 +195,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown,
why don't you just have shutdown use the same exact function as remove? Frankly, though, I still don't fully understand what's going wrong here. Why is the device still alive during kexec?
cheers
Signature · 1 line
balbi
_______________________________________________
Linux-rockchip mailing list [email protected]
http://lists.infradead.org/mailman/listinfo/linux-rockchip
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Quoted 6 lines · Felipe Balbi
Hi,
Quoted 3 lines · Vicente Bergas
Quoted 2 lines · Vicente Bergas
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
...
Quoted 6 lines
why don't you just have shutdown use the same exact function as remove? Frankly, though, I still don't fully understand what's going wrong here. Why is the device still alive during kexec?
cheers
Hi Felipe,
the remove and shutdown functions have different prototypes, so shutdown is wrapping remove.
Would it be preferable to cast remove as shutdown?
The issue with kexec is that the device is being used during the livetime of the first kernel. When the first kernel executes kexec it calls the shutdown function of drivers (instead of remove). Because of this the dwc3 device keeps doing things like DMA.
While the second kernel is taking over, it gets its memory corrupted with such DMA accesses from the device. When the second kernel reaches the point of taking over the dwc3 device, re-initializes it, but it is already too late. Still worse, if the second kernel did not have the dwc3 driver, it would get endless memory corruptions.
All in all, devices that can do DMA need to stop doing it on shutdown.
Regards,
Vicenç.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Quoted 31 lines · Vicente Bergas
Quoted 6 lines · Felipe Balbi
Hi,
Quoted 3 lines · Vicente Bergas
Quoted 2 lines · Vicente Bergas
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
...
Quoted 6 lines
why don't you just have shutdown use the same exact function as remove? Frankly, though, I still don't fully understand what's going wrong here. Why is the device still alive during kexec?
cheers
Hi Felipe,
the remove and shutdown functions have different prototypes, so shutdown is wrapping remove.
Would it be preferable to cast remove as shutdown?The issue with kexec is that the device is being used during the livetime of the first kernel. When the first kernel executes kexec it calls the shutdown function of drivers (instead of remove). Because of this the dwc3 device keeps doing things like DMA.
While the second kernel is taking over, it gets its memory corrupted with such DMA accesses from the device. When the second kernel reaches the point of taking over the dwc3 device, re-initializes it, but it is already too late. Still worse, if the second kernel did not have the dwc3 driver, it would get endless memory corruptions.
All in all, devices that can do DMA need to stop doing it on shutdown.
Quoted 2 lines
Regards,
Vicenç.
Hi,
please, can you provide some feedback on this?
Regards,
Vicenç.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Hi,
(sorry for the long delay)
Quoted 32 lines · Vicente Bergas
Quoted 31 lines · Vicente Bergas
Quoted 6 lines · Felipe Balbi
Hi,
Quoted 3 lines · Vicente Bergas
Quoted 2 lines · Vicente Bergas
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
...
Quoted 6 lines
why don't you just have shutdown use the same exact function as remove? Frankly, though, I still don't fully understand what's going wrong here. Why is the device still alive during kexec?
cheers
Hi Felipe,
the remove and shutdown functions have different prototypes, so shutdown is wrapping remove.
Would it be preferable to cast remove as shutdown?The issue with kexec is that the device is being used during the livetime of the first kernel. When the first kernel executes kexec it calls the shutdown function of drivers (instead of remove). Because of this the dwc3 device keeps doing things like DMA.
While the second kernel is taking over, it gets its memory corrupted with such DMA accesses from the device. When the second kernel reaches the point of taking over the dwc3 device, re-initializes it, but it is already too late. Still worse, if the second kernel did not have the dwc3 driver, it would get endless memory corruptions.
All in all, devices that can do DMA need to stop doing it on shutdown.
Quoted 5 lines
Quoted 2 lines
Regards,
Vicenç.Hi,
please, can you provide some feedback on this?
I meant something like this:
drivers/usb/dwc3/dwc3-of-simple.c +20 -8
@@ -110,12 +110,9 @@ static int dwc3_of_simple_probe(struct platform_device *pdev)return ret; } -static int dwc3_of_simple_remove (struct platform_device *pdev) +static void __dwc3_of_simple_teardown (struct dwc3_of_simple *simple) { -struct dwc3_of_simple *simple =platform_get_drvdata (pdev); -struct device *dev = &pdev->dev; - -of_platform_depopulate (dev); +of_platform_depopulate (simple->dev);clk_bulk_disable_unprepare (simple->num_clocks, simple->clks);clk_bulk_put_all (simple->num_clocks, simple->clks);
@@ -126,13 +123,27 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)reset_control_put (simple->resets); -pm_runtime_disable (dev); -pm_runtime_put_noidle (dev); -pm_runtime_set_suspended (dev); +pm_runtime_disable (simple->dev); +pm_runtime_put_noidle (simple->dev); +pm_runtime_set_suspended (simple->dev); +} + +static int dwc3_of_simple_remove (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +__dwc3_of_simple_teardown (simple);return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +__dwc3_of_simple_teardown (simple); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev) {struct dwc3_of_simple *simple =dev_get_drvdata (dev);
@@ -190,6 +201,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
Can you make sure it works as you intended?
Signature · 1 line
balbi
_______________________________________________
Linux-rockchip mailing list [email protected]
http://lists.infradead.org/mailman/listinfo/linux-rockchip
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Quoted 72 lines · Felipe Balbi
Hi,
(sorry for the long delay)
Quoted 1 line · Vicente Bergas
On Tuesday, August 27, 2019 2:16:20 PM CEST, Vicente Bergas wrote: ...
I meant something like this:
drivers/usb/dwc3/dwc3-of-simple.c +2
-5@@ -110,12 +110,9 @@ static int dwc3_of_simple_probe(struct platform_device *pdev)return ret; } -static int dwc3_of_simple_remove (struct platform_device *pdev) +static void __dwc3_of_simple_teardown (struct dwc3_of_simple *simple) { -struct dwc3_of_simple *simple =platform_get_drvdata (pdev); -struct device *dev = &pdev->dev; - -of_platform_depopulate (dev); +of_platform_depopulate (simple->dev);clk_bulk_disable_unprepare (simple->num_clocks, simple->clks);drivers/usb/dwc3/dwc3-of-simple.c +17
-3clk_bulk_put_all (simple->num_clocks, simple->clks);@@ -126,13 +123,27 @@ static int dwc3_of_simple_remove(struct platform_device *pdev) reset_control_put(simple->resets); -pm_runtime_disable (dev); -pm_runtime_put_noidle (dev); -pm_runtime_set_suspended (dev); +pm_runtime_disable (simple->dev); +pm_runtime_put_noidle (simple->dev); +pm_runtime_set_suspended (simple->dev); +} + +static int dwc3_of_simple_remove (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +__dwc3_of_simple_teardown (simple);return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +__dwc3_of_simple_teardown (simple); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev){
drivers/usb/dwc3/dwc3-of-simple.c +1
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);@@ -190,6 +201,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
Quoted 1 line
Can you make sure it works as you intended?
Hi Felipe,
just applied your approach to v5.3.7 and it is working properly.
Thanks,
Vicente.
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
hi,
Quoted 74 lines · Vicente Bergas
Quoted 73 lines · Felipe Balbi
Hi,
(sorry for the long delay)
Quoted 1 line · Vicente Bergas
On Tuesday, August 27, 2019 2:16:20 PM CEST, Vicente Bergas wrote: ...
I meant something like this:
drivers/usb/dwc3/dwc3-of-simple.c +2
-5@@ -110,12 +110,9 @@ static int dwc3_of_simple_probe(struct platform_device *pdev)return ret; } -static int dwc3_of_simple_remove (struct platform_device *pdev) +static void __dwc3_of_simple_teardown (struct dwc3_of_simple *simple) { -struct dwc3_of_simple *simple =platform_get_drvdata (pdev); -struct device *dev = &pdev->dev; - -of_platform_depopulate (dev); +of_platform_depopulate (simple->dev);clk_bulk_disable_unprepare (simple->num_clocks, simple->clks);drivers/usb/dwc3/dwc3-of-simple.c +17
-3clk_bulk_put_all (simple->num_clocks, simple->clks);@@ -126,13 +123,27 @@ static int dwc3_of_simple_remove(struct platform_device *pdev) reset_control_put(simple->resets); -pm_runtime_disable (dev); -pm_runtime_put_noidle (dev); -pm_runtime_set_suspended (dev); +pm_runtime_disable (simple->dev); +pm_runtime_put_noidle (simple->dev); +pm_runtime_set_suspended (simple->dev); +} + +static int dwc3_of_simple_remove (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +__dwc3_of_simple_teardown (simple);return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +struct dwc3_of_simple *simple =platform_get_drvdata (pdev); + +__dwc3_of_simple_teardown (simple); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev){
drivers/usb/dwc3/dwc3-of-simple.c +1
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);@@ -190,6 +201,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
Quoted 4 lines
Quoted 1 line
Can you make sure it works as you intended?
Hi Felipe,
just applied your approach to v5.3.7 and it is working properly.
Do you want to send it as a formal patch or shall I do it?
Signature · 1 line
balbi
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
Quoted 7 lines · Felipe Balbi
hi,
Quoted 1 line · Vicente Bergas
On Wednesday, October 23, 2019 8:31:21 AM CEST, Felipe Balbi wrote: ...
Do you want to send it as a formal patch or shall I do it?
All yours. Thank you for reviewing and proposing this solution.
Regards,
Vicente.
Re: [PATCH] usb: dwc3: Add shutdown to platform_driver
In reply to Vicente BergasPing?
Quoted 39 lines · Vicente Bergas
Otherwise the device keeps writing to memory after kexec and disturbs the next kernel.
Signed-off-by Vicente Bergas <[email protected]>--- drivers/usb/dwc3/dwc3-of-simple.c | 6 ++++++ 1 file changed, 6 insertions(+)Hi Felipe, Robin,
this version calls 'remove' from 'shutdown' instead of just asserting a reset because it looks like a cleaner way to stop the device.Calling remove from shutdown in
core.cinstead ofdwc3-of-simple.cdoes not fix the issue either.It has been tested on the sapphire board, a RK3399 platform.
Regards,
Vicenç.drivers/usb/dwc3/dwc3-of-simple.c +5
-0@@ -133,6 +133,11 @@ static int dwc3_of_simple_remove(struct platform_device *pdev)return 0 ; } +static void dwc3_of_simple_shutdown (struct platform_device *pdev) +{ +dwc3_of_simple_remove (pdev); +} +static int __maybe_unused dwc3_of_simple_runtime_suspend (struct device *dev){
drivers/usb/dwc3/dwc3-of-simple.c +0
-0struct dwc3_of_simple *simple =dev_get_drvdata (dev);
drivers/usb/dwc3/dwc3-of-simple.c MODULE_DEVICE_TABLE(of, of_dwc3_simple_match); · line 201
Quoted 7 lines
drivers/usb/dwc3/dwc3-of-simple.c +1
-0@@ -190,6 +195,7 @@ MODULE_DEVICE_TABLE(of, of_dwc3_simple_match);static struct platform_driver dwc3_of_simple_driver = { .probe = dwc3_of_simple_probe, .remove = dwc3_of_simple_remove, + .shutdown = dwc3_of_simple_shutdown, .driver = { .name ="dwc3-of-simple" , .of_match_table = of_dwc3_simple_match,
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel