[PATCH 040/190] Revert "ACPI: CPPC: Fix reference count leak in acpi_cppc_processor_probe()"

[PATCH 040/190] Revert "ACPI: CPPC: Fix reference count leak in acpi_cppc_processor_probe()"

[Greg Kroah-Hartman]

This reverts commit 4d8be4bc94f74bb7d096e1c2e44457b530d5a170.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: Qiushi Wu <wu000273@umn.edu>
Cc: 4.10+ <stable@vger.kernel.org> # 4.10+
Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/acpi/cppc_acpi.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/acpi/cppc_acpi.c b/drivers/acpi/cppc_acpi.c
index 69057fcd2c04..42650b34e45e 100644
--- a/drivers/acpi/cppc_acpi.c
+++ b/drivers/acpi/cppc_acpi.c
@@ -830,7 +830,6 @@ int acpi_cppc_processor_probe(struct acpi_processor *pr)
 			"acpi_cppc");
 	if (ret) {
 		per_cpu(cpc_desc_ptr, pr->id) = NULL;
-		kobject_put(&cpc_ptr->kobj);
 		goto out_free;
 	}
 
-- 
2.31.1

Re: [PATCH 040/190] Revert "ACPI: CPPC: Fix reference count leak in acpi_cppc_processor_probe()"

[Rafael J. Wysocki]

On 4/21/2021 2:58 PM, Greg Kroah-Hartman wrote:
> This reverts commit 4d8be4bc94f74bb7d096e1c2e44457b530d5a170.
>
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
>
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
>
> Cc: Qiushi Wu <wu000273@umn.edu>
> Cc: 4.10+ <stable@vger.kernel.org> # 4.10+
> Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>


> ---
>   drivers/acpi/cppc_acpi.c | 1 -
>   1 file changed, 1 deletion(-)
>
> diff --git a/drivers/acpi/cppc_acpi.c b/drivers/acpi/cppc_acpi.c
> index 69057fcd2c04..42650b34e45e 100644
> --- a/drivers/acpi/cppc_acpi.c
> +++ b/drivers/acpi/cppc_acpi.c
> @@ -830,7 +830,6 @@ int acpi_cppc_processor_probe(struct acpi_processor *pr)
>   			"acpi_cppc");
>   	if (ret) {
>   		per_cpu(cpc_desc_ptr, pr->id) = NULL;
> -		kobject_put(&cpc_ptr->kobj);
>   		goto out_free;
>   	}
>   

Re: [PATCH 040/190] Revert "ACPI: CPPC: Fix reference count leak in acpi_cppc_processor_probe()"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 02:58:35PM +0200, Greg Kroah-Hartman wrote:
> This reverts commit 4d8be4bc94f74bb7d096e1c2e44457b530d5a170.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: Qiushi Wu <wu000273@umn.edu>
> Cc: 4.10+ <stable@vger.kernel.org> # 4.10+
> Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>  drivers/acpi/cppc_acpi.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/acpi/cppc_acpi.c b/drivers/acpi/cppc_acpi.c
> index 69057fcd2c04..42650b34e45e 100644
> --- a/drivers/acpi/cppc_acpi.c
> +++ b/drivers/acpi/cppc_acpi.c
> @@ -830,7 +830,6 @@ int acpi_cppc_processor_probe(struct acpi_processor *pr)
>  			"acpi_cppc");
>  	if (ret) {
>  		per_cpu(cpc_desc_ptr, pr->id) = NULL;
> -		kobject_put(&cpc_ptr->kobj);
>  		goto out_free;
>  	}
>  
> -- 
> 2.31.1
>

The original change here looks correct, so I will drop this revert from
my tree.

greg k-h

[PATCH 041/190] Revert "ACPI: sysfs: Fix reference count leak in acpi_sysfs_add_hotplug_profile()"

[Greg Kroah-Hartman]

This reverts commit 6e6c25283dff866308c87b49434c7dbad4774cc0.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: Qiushi Wu <wu000273@umn.edu>
Cc: 3.10+ <stable@vger.kernel.org> # 3.10+
Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/acpi/sysfs.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/acpi/sysfs.c b/drivers/acpi/sysfs.c
index 8baf7644a0d0..842bf63b91e9 100644
--- a/drivers/acpi/sysfs.c
+++ b/drivers/acpi/sysfs.c
@@ -986,10 +986,8 @@ void acpi_sysfs_add_hotplug_profile(struct acpi_hotplug_profile *hotplug,
 
 	error = kobject_init_and_add(&hotplug->kobj,
 		&acpi_hotplug_profile_ktype, hotplug_kobj, "%s", name);
-	if (error) {
-		kobject_put(&hotplug->kobj);
+	if (error)
 		goto err_out;
-	}
 
 	kobject_uevent(&hotplug->kobj, KOBJ_ADD);
 	return;
-- 
2.31.1

Re: [PATCH 041/190] Revert "ACPI: sysfs: Fix reference count leak in acpi_sysfs_add_hotplug_profile()"

[Rafael J. Wysocki]

On 4/21/2021 2:58 PM, Greg Kroah-Hartman wrote:
> This reverts commit 6e6c25283dff866308c87b49434c7dbad4774cc0.
>
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
>
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
>
> Cc: Qiushi Wu <wu000273@umn.edu>
> Cc: 3.10+ <stable@vger.kernel.org> # 3.10+
> Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>


> ---
>   drivers/acpi/sysfs.c | 4 +---
>   1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/acpi/sysfs.c b/drivers/acpi/sysfs.c
> index 8baf7644a0d0..842bf63b91e9 100644
> --- a/drivers/acpi/sysfs.c
> +++ b/drivers/acpi/sysfs.c
> @@ -986,10 +986,8 @@ void acpi_sysfs_add_hotplug_profile(struct acpi_hotplug_profile *hotplug,
>   
>   	error = kobject_init_and_add(&hotplug->kobj,
>   		&acpi_hotplug_profile_ktype, hotplug_kobj, "%s", name);
> -	if (error) {
> -		kobject_put(&hotplug->kobj);
> +	if (error)
>   		goto err_out;
> -	}
>   
>   	kobject_uevent(&hotplug->kobj, KOBJ_ADD);
>   	return;

Re: [PATCH 041/190] Revert "ACPI: sysfs: Fix reference count leak in acpi_sysfs_add_hotplug_profile()"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 07:54:23PM +0200, Rafael J. Wysocki wrote:
> On 4/21/2021 2:58 PM, Greg Kroah-Hartman wrote:
> > This reverts commit 6e6c25283dff866308c87b49434c7dbad4774cc0.
> > 
> > Commits from @umn.edu addresses have been found to be submitted in "bad
> > faith" to try to test the kernel community's ability to review "known
> > malicious" changes.  The result of these submissions can be found in a
> > paper published at the 42nd IEEE Symposium on Security and Privacy
> > entitled, "Open Source Insecurity: Stealthily Introducing
> > Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> > of Minnesota) and Kangjie Lu (University of Minnesota).
> > 
> > Because of this, all submissions from this group must be reverted from
> > the kernel tree and will need to be re-reviewed again to determine if
> > they actually are a valid fix.  Until that work is complete, remove this
> > change to ensure that no problems are being introduced into the
> > codebase.
> > 
> > Cc: Qiushi Wu <wu000273@umn.edu>
> > Cc: 3.10+ <stable@vger.kernel.org> # 3.10+
> > Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> > Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> 
> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Thanks for the review (and on the other acpi patch), but this commit
looks like it was correct, so I am going to drop the revert from my
tree.

greg k-h

[PATCH 081/190] Revert "tracing: Fix a memory leak by early error exit in trace_pid_write()"

[Greg Kroah-Hartman]

This reverts commit 91862cc7867bba4ee5c8fcf0ca2f1d30427b6129.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: http
Cc: stable@vger.kernel.org
Cc: Wenwen Wang <wang6495@umn.edu>
Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 kernel/trace/trace.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 5c777627212f..faed4f44d224 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -691,10 +691,8 @@ int trace_pid_write(struct trace_pid_list *filtered_pids,
 	 * not modified.
 	 */
 	pid_list = kmalloc(sizeof(*pid_list), GFP_KERNEL);
-	if (!pid_list) {
-		trace_parser_put(&parser);
+	if (!pid_list)
 		return -ENOMEM;
-	}
 
 	pid_list->pid_max = READ_ONCE(pid_max);
 
@@ -704,7 +702,6 @@ int trace_pid_write(struct trace_pid_list *filtered_pids,
 
 	pid_list->pids = vzalloc((pid_list->pid_max + 7) >> 3);
 	if (!pid_list->pids) {
-		trace_parser_put(&parser);
 		kfree(pid_list);
 		return -ENOMEM;
 	}
-- 
2.31.1

Re: [PATCH 081/190] Revert "tracing: Fix a memory leak by early error exit in trace_pid_write()"

[Steven Rostedt]

On Wed, 21 Apr 2021 14:59:16 +0200
Greg Kroah-Hartman <gregkh@linuxfoundation.org> wrote:

> This reverts commit 91862cc7867bba4ee5c8fcf0ca2f1d30427b6129.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 

I have reviewed this change, and this is a valid fix and does not need to
be reverted.

The code before the change is:

	if (trace_parser_get_init(&parser, PID_BUF_SIZE + 1))
		return -ENOMEM;

Where that does:

int trace_parser_get_init(struct trace_parser *parser, int size)
{
	memset(parser, 0, sizeof(*parser));

	parser->buffer = kmalloc(size, GFP_KERNEL);
	if (!parser->buffer)
		return 1;

	parser->size = size;
	return 0;
}

And the trace_parser_put() does:

void trace_parser_put(struct trace_parser *parser)
{
	kfree(parser->buffer);
	parser->buffer = NULL;
}

Hence, exiting the function without calling trace_parser_put() will indeed
leak memory.

Please do not revert this patch.

Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>

-- Steve


> Cc: http
> Cc: stable@vger.kernel.org
> Cc: Wenwen Wang <wang6495@umn.edu>
> Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>  kernel/trace/trace.c | 5 +----
>  1 file changed, 1 insertion(+), 4 deletions(-)
> 
> diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> index 5c777627212f..faed4f44d224 100644
> --- a/kernel/trace/trace.c
> +++ b/kernel/trace/trace.c
> @@ -691,10 +691,8 @@ int trace_pid_write(struct trace_pid_list *filtered_pids,
>  	 * not modified.
>  	 */
>  	pid_list = kmalloc(sizeof(*pid_list), GFP_KERNEL);
> -	if (!pid_list) {
> -		trace_parser_put(&parser);
> +	if (!pid_list)
>  		return -ENOMEM;
> -	}
>  
>  	pid_list->pid_max = READ_ONCE(pid_max);
>  
> @@ -704,7 +702,6 @@ int trace_pid_write(struct trace_pid_list *filtered_pids,
>  
>  	pid_list->pids = vzalloc((pid_list->pid_max + 7) >> 3);
>  	if (!pid_list->pids) {
> -		trace_parser_put(&parser);
>  		kfree(pid_list);
>  		return -ENOMEM;
>  	}

Re: [PATCH 081/190] Revert "tracing: Fix a memory leak by early error exit in trace_pid_write()"

[Steven Rostedt]

On Wed, 21 Apr 2021 09:29:19 -0400
Steven Rostedt <rostedt@goodmis.org> wrote:

> Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>

Just to clear up any confusion about my tag above. It was a second review
of the original patch, not for the revert.

-- Steve

Re: [PATCH 081/190] Revert "tracing: Fix a memory leak by early error exit in trace_pid_write()"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 09:33:43AM -0400, Steven Rostedt wrote:
> On Wed, 21 Apr 2021 09:29:19 -0400
> Steven Rostedt <rostedt@goodmis.org> wrote:
> 
> > Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
> 
> Just to clear up any confusion about my tag above. It was a second review
> of the original patch, not for the revert.

Fair enough, I'll handle it, thanks!

greg k-h

Re: [PATCH 081/190] Revert "tracing: Fix a memory leak by early error exit in trace_pid_write()"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 03:51:59PM +0200, Greg Kroah-Hartman wrote:
> On Wed, Apr 21, 2021 at 09:33:43AM -0400, Steven Rostedt wrote:
> > On Wed, 21 Apr 2021 09:29:19 -0400
> > Steven Rostedt <rostedt@goodmis.org> wrote:
> > 
> > > Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
> > 
> > Just to clear up any confusion about my tag above. It was a second review
> > of the original patch, not for the revert.
> 
> Fair enough, I'll handle it, thanks!

Revert is now dropped from my tree, thanks for the review.

greg k-h

[PATCH 119/190] Revert "tty: mxs-auart: fix a potential NULL pointer dereference"

[Greg Kroah-Hartman]

This reverts commit 6734330654dac550f12e932996b868c6d0dcb421.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: Kangjie Lu <kjlu@umn.edu>
Cc: stable <stable@vger.kernel.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/tty/serial/mxs-auart.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/drivers/tty/serial/mxs-auart.c b/drivers/tty/serial/mxs-auart.c
index f414d6acad69..edad6ebbdfd5 100644
--- a/drivers/tty/serial/mxs-auart.c
+++ b/drivers/tty/serial/mxs-auart.c
@@ -1644,10 +1644,6 @@ static int mxs_auart_probe(struct platform_device *pdev)
 
 	s->port.mapbase = r->start;
 	s->port.membase = ioremap(r->start, resource_size(r));
-	if (!s->port.membase) {
-		ret = -ENOMEM;
-		goto out_disable_clks;
-	}
 	s->port.ops = &mxs_auart_ops;
 	s->port.iotype = UPIO_MEM;
 	s->port.fifosize = MXS_AUART_FIFO_SIZE;
-- 
2.31.1

Re: [PATCH 119/190] Revert "tty: mxs-auart: fix a potential NULL pointer dereference"

[Jiri Slaby]

On 21. 04. 21, 14:59, Greg Kroah-Hartman wrote:
> This reverts commit 6734330654dac550f12e932996b868c6d0dcb421.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: Kangjie Lu <kjlu@umn.edu>
> Cc: stable <stable@vger.kernel.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>   drivers/tty/serial/mxs-auart.c | 4 ----
>   1 file changed, 4 deletions(-)
> 
> diff --git a/drivers/tty/serial/mxs-auart.c b/drivers/tty/serial/mxs-auart.c
> index f414d6acad69..edad6ebbdfd5 100644
> --- a/drivers/tty/serial/mxs-auart.c
> +++ b/drivers/tty/serial/mxs-auart.c
> @@ -1644,10 +1644,6 @@ static int mxs_auart_probe(struct platform_device *pdev)
>   
>   	s->port.mapbase = r->start;
>   	s->port.membase = ioremap(r->start, resource_size(r));
> -	if (!s->port.membase) {
> -		ret = -ENOMEM;
> -		goto out_disable_clks;
> -	}

I don't think this needs to be reverted -- the original fix is correct.

>   	s->port.ops = &mxs_auart_ops;
>   	s->port.iotype = UPIO_MEM;
>   	s->port.fifosize = MXS_AUART_FIFO_SIZE;
> 


-- 
js
suse labs

Re: [PATCH 119/190] Revert "tty: mxs-auart: fix a potential NULL pointer dereference"

[Greg Kroah-Hartman]

On Thu, Apr 22, 2021 at 07:03:55AM +0200, Jiri Slaby wrote:
> On 21. 04. 21, 14:59, Greg Kroah-Hartman wrote:
> > This reverts commit 6734330654dac550f12e932996b868c6d0dcb421.
> > 
> > Commits from @umn.edu addresses have been found to be submitted in "bad
> > faith" to try to test the kernel community's ability to review "known
> > malicious" changes.  The result of these submissions can be found in a
> > paper published at the 42nd IEEE Symposium on Security and Privacy
> > entitled, "Open Source Insecurity: Stealthily Introducing
> > Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> > of Minnesota) and Kangjie Lu (University of Minnesota).
> > 
> > Because of this, all submissions from this group must be reverted from
> > the kernel tree and will need to be re-reviewed again to determine if
> > they actually are a valid fix.  Until that work is complete, remove this
> > change to ensure that no problems are being introduced into the
> > codebase.
> > 
> > Cc: Kangjie Lu <kjlu@umn.edu>
> > Cc: stable <stable@vger.kernel.org>
> > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > ---
> >   drivers/tty/serial/mxs-auart.c | 4 ----
> >   1 file changed, 4 deletions(-)
> > 
> > diff --git a/drivers/tty/serial/mxs-auart.c b/drivers/tty/serial/mxs-auart.c
> > index f414d6acad69..edad6ebbdfd5 100644
> > --- a/drivers/tty/serial/mxs-auart.c
> > +++ b/drivers/tty/serial/mxs-auart.c
> > @@ -1644,10 +1644,6 @@ static int mxs_auart_probe(struct platform_device *pdev)
> >   	s->port.mapbase = r->start;
> >   	s->port.membase = ioremap(r->start, resource_size(r));
> > -	if (!s->port.membase) {
> > -		ret = -ENOMEM;
> > -		goto out_disable_clks;
> > -	}
> 
> I don't think this needs to be reverted -- the original fix is correct.
> 

Now dropped, thanks for the review!

greg k-h

[PATCH 120/190] Revert "tty: atmel_serial: fix a potential NULL pointer dereference"

[Greg Kroah-Hartman]

This reverts commit c85be041065c0be8bc48eda4c45e0319caf1d0e5.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: Kangjie Lu <kjlu@umn.edu>
Cc: Richard Genoud <richard.genoud@gmail.com>
Cc: stable <stable@vger.kernel.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/tty/serial/atmel_serial.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/drivers/tty/serial/atmel_serial.c b/drivers/tty/serial/atmel_serial.c
index a24e5c2b30bc..9786d8e5f04f 100644
--- a/drivers/tty/serial/atmel_serial.c
+++ b/drivers/tty/serial/atmel_serial.c
@@ -1256,10 +1256,6 @@ static int atmel_prepare_rx_dma(struct uart_port *port)
 					 sg_dma_len(&atmel_port->sg_rx)/2,
 					 DMA_DEV_TO_MEM,
 					 DMA_PREP_INTERRUPT);
-	if (!desc) {
-		dev_err(port->dev, "Preparing DMA cyclic failed\n");
-		goto chan_err;
-	}
 	desc->callback = atmel_complete_rx_dma;
 	desc->callback_param = port;
 	atmel_port->desc_rx = desc;
-- 
2.31.1

Re: [PATCH 120/190] Revert "tty: atmel_serial: fix a potential NULL pointer dereference"

[Jiri Slaby]

On 21. 04. 21, 14:59, Greg Kroah-Hartman wrote:
> This reverts commit c85be041065c0be8bc48eda4c45e0319caf1d0e5.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: Kangjie Lu <kjlu@umn.edu>
> Cc: Richard Genoud <richard.genoud@gmail.com>
> Cc: stable <stable@vger.kernel.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>   drivers/tty/serial/atmel_serial.c | 4 ----
>   1 file changed, 4 deletions(-)
> 
> diff --git a/drivers/tty/serial/atmel_serial.c b/drivers/tty/serial/atmel_serial.c
> index a24e5c2b30bc..9786d8e5f04f 100644
> --- a/drivers/tty/serial/atmel_serial.c
> +++ b/drivers/tty/serial/atmel_serial.c
> @@ -1256,10 +1256,6 @@ static int atmel_prepare_rx_dma(struct uart_port *port)
>   					 sg_dma_len(&atmel_port->sg_rx)/2,
>   					 DMA_DEV_TO_MEM,
>   					 DMA_PREP_INTERRUPT);
> -	if (!desc) {
> -		dev_err(port->dev, "Preparing DMA cyclic failed\n");
> -		goto chan_err;
> -	}

I cannot find anything malicious in the original fix:
* port->dev is valid for dev_err
* dmaengine_prep_dma_cyclic returns NULL in case of error
* chan_err invokes atmel_release_rx_dma which undoes the previous 
initialization code.

Hence a NACK from me for the revert.

>   	desc->callback = atmel_complete_rx_dma;
>   	desc->callback_param = port;
>   	atmel_port->desc_rx = desc;
> 


-- 
js
suse labs

Re: [PATCH 120/190] Revert "tty: atmel_serial: fix a potential NULL pointer dereference"

[Richard Genoud]

Le 22/04/2021 à 07:18, Jiri Slaby a écrit :
> On 21. 04. 21, 14:59, Greg Kroah-Hartman wrote:
>> This reverts commit c85be041065c0be8bc48eda4c45e0319caf1d0e5.
>>
>> Commits from @umn.edu addresses have been found to be submitted in "bad
>> faith" to try to test the kernel community's ability to review "known
>> malicious" changes.  The result of these submissions can be found in a
>> paper published at the 42nd IEEE Symposium on Security and Privacy
>> entitled, "Open Source Insecurity: Stealthily Introducing
>> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
>> of Minnesota) and Kangjie Lu (University of Minnesota).
>>
>> Because of this, all submissions from this group must be reverted from
>> the kernel tree and will need to be re-reviewed again to determine if
>> they actually are a valid fix.  Until that work is complete, remove this
>> change to ensure that no problems are being introduced into the
>> codebase.
>>
>> Cc: Kangjie Lu <kjlu@umn.edu>
>> Cc: Richard Genoud <richard.genoud@gmail.com>
>> Cc: stable <stable@vger.kernel.org>
>> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> ---
>>   drivers/tty/serial/atmel_serial.c | 4 ----
>>   1 file changed, 4 deletions(-)
>>
>> diff --git a/drivers/tty/serial/atmel_serial.c b/drivers/tty/serial/atmel_serial.c
>> index a24e5c2b30bc..9786d8e5f04f 100644
>> --- a/drivers/tty/serial/atmel_serial.c
>> +++ b/drivers/tty/serial/atmel_serial.c
>> @@ -1256,10 +1256,6 @@ static int atmel_prepare_rx_dma(struct uart_port *port)
>>                        sg_dma_len(&atmel_port->sg_rx)/2,
>>                        DMA_DEV_TO_MEM,
>>                        DMA_PREP_INTERRUPT);
>> -    if (!desc) {
>> -        dev_err(port->dev, "Preparing DMA cyclic failed\n");
>> -        goto chan_err;
>> -    }
> 
> I cannot find anything malicious in the original fix:
> * port->dev is valid for dev_err
> * dmaengine_prep_dma_cyclic returns NULL in case of error
> * chan_err invokes atmel_release_rx_dma which undoes the previous initialization code.
> 
> Hence a NACK from me for the revert.

I agree with your NACK.
Back at the time (march 2019), I reviewed the changed and asked for a 2nd version and
I didn't found anything suspicious.
But the more eyes, the better.

cf http://lkml.iu.edu/hypermail/linux/kernel/1903.1/05858.html

> 
>>       desc->callback = atmel_complete_rx_dma;
>>       desc->callback_param = port;
>>       atmel_port->desc_rx = desc;
>>
> 
> 

Re: [PATCH 120/190] Revert "tty: atmel_serial: fix a potential NULL pointer dereference"

[Greg Kroah-Hartman]

On Thu, Apr 22, 2021 at 06:47:20AM +0000, Richard Genoud wrote:
> 
> 
> Le 22/04/2021 à 07:18, Jiri Slaby a écrit :
> > On 21. 04. 21, 14:59, Greg Kroah-Hartman wrote:
> > > This reverts commit c85be041065c0be8bc48eda4c45e0319caf1d0e5.
> > > 
> > > Commits from @umn.edu addresses have been found to be submitted in "bad
> > > faith" to try to test the kernel community's ability to review "known
> > > malicious" changes.  The result of these submissions can be found in a
> > > paper published at the 42nd IEEE Symposium on Security and Privacy
> > > entitled, "Open Source Insecurity: Stealthily Introducing
> > > Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> > > of Minnesota) and Kangjie Lu (University of Minnesota).
> > > 
> > > Because of this, all submissions from this group must be reverted from
> > > the kernel tree and will need to be re-reviewed again to determine if
> > > they actually are a valid fix.  Until that work is complete, remove this
> > > change to ensure that no problems are being introduced into the
> > > codebase.
> > > 
> > > Cc: Kangjie Lu <kjlu@umn.edu>
> > > Cc: Richard Genoud <richard.genoud@gmail.com>
> > > Cc: stable <stable@vger.kernel.org>
> > > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > ---
> > >   drivers/tty/serial/atmel_serial.c | 4 ----
> > >   1 file changed, 4 deletions(-)
> > > 
> > > diff --git a/drivers/tty/serial/atmel_serial.c b/drivers/tty/serial/atmel_serial.c
> > > index a24e5c2b30bc..9786d8e5f04f 100644
> > > --- a/drivers/tty/serial/atmel_serial.c
> > > +++ b/drivers/tty/serial/atmel_serial.c
> > > @@ -1256,10 +1256,6 @@ static int atmel_prepare_rx_dma(struct uart_port *port)
> > >                        sg_dma_len(&atmel_port->sg_rx)/2,
> > >                        DMA_DEV_TO_MEM,
> > >                        DMA_PREP_INTERRUPT);
> > > -    if (!desc) {
> > > -        dev_err(port->dev, "Preparing DMA cyclic failed\n");
> > > -        goto chan_err;
> > > -    }
> > 
> > I cannot find anything malicious in the original fix:
> > * port->dev is valid for dev_err
> > * dmaengine_prep_dma_cyclic returns NULL in case of error
> > * chan_err invokes atmel_release_rx_dma which undoes the previous initialization code.
> > 
> > Hence a NACK from me for the revert.
> 
> I agree with your NACK.
> Back at the time (march 2019), I reviewed the changed and asked for a 2nd version and
> I didn't found anything suspicious.
> But the more eyes, the better.
> 
> cf http://lkml.iu.edu/hypermail/linux/kernel/1903.1/05858.html

Thanks for the review, now dropped.

greg k-h

[PATCH 121/190] Revert "serial: mvebu-uart: Fix to avoid a potential NULL pointer dereference"

[Greg Kroah-Hartman]

This reverts commit 32f47179833b63de72427131169809065db6745e.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: Aditya Pakki <pakki001@umn.edu>
Cc: stable <stable@vger.kernel.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/tty/serial/mvebu-uart.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/tty/serial/mvebu-uart.c b/drivers/tty/serial/mvebu-uart.c
index e0c00a1b0763..51b0ecabf2ec 100644
--- a/drivers/tty/serial/mvebu-uart.c
+++ b/drivers/tty/serial/mvebu-uart.c
@@ -818,9 +818,6 @@ static int mvebu_uart_probe(struct platform_device *pdev)
 		return -EINVAL;
 	}
 
-	if (!match)
-		return -ENODEV;
-
 	/* Assume that all UART ports have a DT alias or none has */
 	id = of_alias_get_id(pdev->dev.of_node, "serial");
 	if (!pdev->dev.of_node || id < 0)
-- 
2.31.1

Re: [PATCH 121/190] Revert "serial: mvebu-uart: Fix to avoid a potential NULL pointer dereference"

[Jiri Slaby]

On 21. 04. 21, 14:59, Greg Kroah-Hartman wrote:
> This reverts commit 32f47179833b63de72427131169809065db6745e.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: Aditya Pakki <pakki001@umn.edu>
> Cc: stable <stable@vger.kernel.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>   drivers/tty/serial/mvebu-uart.c | 3 ---
>   1 file changed, 3 deletions(-)
> 
> diff --git a/drivers/tty/serial/mvebu-uart.c b/drivers/tty/serial/mvebu-uart.c
> index e0c00a1b0763..51b0ecabf2ec 100644
> --- a/drivers/tty/serial/mvebu-uart.c
> +++ b/drivers/tty/serial/mvebu-uart.c
> @@ -818,9 +818,6 @@ static int mvebu_uart_probe(struct platform_device *pdev)
>   		return -EINVAL;
>   	}
>   
> -	if (!match)
> -		return -ENODEV;

The original fix doesn't hurt, but is useless. ->probe is called when 
of_match_device matched.

So the revert is OK.

>   	/* Assume that all UART ports have a DT alias or none has */
>   	id = of_alias_get_id(pdev->dev.of_node, "serial");
>   	if (!pdev->dev.of_node || id < 0)
> 


-- 
js
suse labs

[PATCH 134/190] Revert "md: Fix failed allocation of md_register_thread"

[Greg Kroah-Hartman]

This reverts commit e406f12dde1a8375d77ea02d91f313fb1a9c6aec.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: stable@vger.kernel.org # v3.16+
Cc: Guoqing Jiang <gqjiang@suse.com>
Cc: Aditya Pakki <pakki001@umn.edu>
Cc: Song Liu <songliubraving@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/md/raid10.c | 2 --
 drivers/md/raid5.c  | 2 --
 2 files changed, 4 deletions(-)

diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index a9ae7d113492..4fec1cdd4207 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -3896,8 +3896,6 @@ static int raid10_run(struct mddev *mddev)
 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
 							"reshape");
-		if (!mddev->sync_thread)
-			goto out_free_conf;
 	}
 
 	return 0;
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 5d57a5bd171f..9b2bd50beee7 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -7677,8 +7677,6 @@ static int raid5_run(struct mddev *mddev)
 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
 							"reshape");
-		if (!mddev->sync_thread)
-			goto abort;
 	}
 
 	/* Ok, everything is just fine now */
-- 
2.31.1

Re: [PATCH 134/190] Revert "md: Fix failed allocation of md_register_thread"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 03:00:09PM +0200, Greg Kroah-Hartman wrote:
> This reverts commit e406f12dde1a8375d77ea02d91f313fb1a9c6aec.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: stable@vger.kernel.org # v3.16+
> Cc: Guoqing Jiang <gqjiang@suse.com>
> Cc: Aditya Pakki <pakki001@umn.edu>
> Cc: Song Liu <songliubraving@fb.com>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>  drivers/md/raid10.c | 2 --
>  drivers/md/raid5.c  | 2 --
>  2 files changed, 4 deletions(-)
> 
> diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
> index a9ae7d113492..4fec1cdd4207 100644
> --- a/drivers/md/raid10.c
> +++ b/drivers/md/raid10.c
> @@ -3896,8 +3896,6 @@ static int raid10_run(struct mddev *mddev)
>  		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
>  		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
>  							"reshape");
> -		if (!mddev->sync_thread)
> -			goto out_free_conf;
>  	}
>  
>  	return 0;
> diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
> index 5d57a5bd171f..9b2bd50beee7 100644
> --- a/drivers/md/raid5.c
> +++ b/drivers/md/raid5.c
> @@ -7677,8 +7677,6 @@ static int raid5_run(struct mddev *mddev)
>  		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
>  		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
>  							"reshape");
> -		if (!mddev->sync_thread)
> -			goto abort;
>  	}
>  
>  	/* Ok, everything is just fine now */
> -- 
> 2.31.1
> 

These changes look ok, but the error handling logic seems to be freeing
the incorrect thread, not the one that these functions create.  That's
independant of this change, but seems odd.  If someone cares about it,
it should probably be looked at, or if correct, a comment would be nice
as it's really confusing.

Dropping this revert.

greg k-h

Re: [PATCH 134/190] Revert "md: Fix failed allocation of md_register_thread"

[Song Liu]

> On Apr 27, 2021, at 10:46 PM, Greg Kroah-Hartman <gregkh@linuxfoundation.org> wrote:
> 
> On Wed, Apr 21, 2021 at 03:00:09PM +0200, Greg Kroah-Hartman wrote:
>> This reverts commit e406f12dde1a8375d77ea02d91f313fb1a9c6aec.
>> 
>> Commits from @umn.edu addresses have been found to be submitted in "bad
>> faith" to try to test the kernel community's ability to review "known
>> malicious" changes.  The result of these submissions can be found in a
>> paper published at the 42nd IEEE Symposium on Security and Privacy
>> entitled, "Open Source Insecurity: Stealthily Introducing
>> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
>> of Minnesota) and Kangjie Lu (University of Minnesota).
>> 
>> Because of this, all submissions from this group must be reverted from
>> the kernel tree and will need to be re-reviewed again to determine if
>> they actually are a valid fix.  Until that work is complete, remove this
>> change to ensure that no problems are being introduced into the
>> codebase.
>> 
>> Cc: stable@vger.kernel.org # v3.16+
>> Cc: Guoqing Jiang <gqjiang@suse.com>
>> Cc: Aditya Pakki <pakki001@umn.edu>
>> Cc: Song Liu <songliubraving@fb.com>
>> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> ---
>> drivers/md/raid10.c | 2 --
>> drivers/md/raid5.c  | 2 --
>> 2 files changed, 4 deletions(-)
>> 
>> diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
>> index a9ae7d113492..4fec1cdd4207 100644
>> --- a/drivers/md/raid10.c
>> +++ b/drivers/md/raid10.c
>> @@ -3896,8 +3896,6 @@ static int raid10_run(struct mddev *mddev)
>> 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
>> 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
>> 							"reshape");
>> -		if (!mddev->sync_thread)
>> -			goto out_free_conf;
>> 	}
>> 
>> 	return 0;
>> diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
>> index 5d57a5bd171f..9b2bd50beee7 100644
>> --- a/drivers/md/raid5.c
>> +++ b/drivers/md/raid5.c
>> @@ -7677,8 +7677,6 @@ static int raid5_run(struct mddev *mddev)
>> 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
>> 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
>> 							"reshape");
>> -		if (!mddev->sync_thread)
>> -			goto abort;
>> 	}
>> 
>> 	/* Ok, everything is just fine now */
>> -- 
>> 2.31.1
>> 
> 
> These changes look ok, but the error handling logic seems to be freeing
> the incorrect thread, not the one that these functions create.  That's
> independant of this change, but seems odd.  If someone cares about it,
> it should probably be looked at, or if correct, a comment would be nice
> as it's really confusing.

I don't think this is confusing. raid[5|10]_run() creates two threads:
first mddev->thread, then mddev->sync_thread. If we fail to create the
second thread (sync_thread), we free the first thread (mddev->thread) in 
the error handling logic. 

Thanks,
Song

Re: [PATCH 134/190] Revert "md: Fix failed allocation of md_register_thread"

[Greg Kroah-Hartman]

On Fri, Apr 30, 2021 at 06:10:24AM +0000, Song Liu wrote:
> 
> 
> > On Apr 27, 2021, at 10:46 PM, Greg Kroah-Hartman <gregkh@linuxfoundation.org> wrote:
> > 
> > On Wed, Apr 21, 2021 at 03:00:09PM +0200, Greg Kroah-Hartman wrote:
> >> This reverts commit e406f12dde1a8375d77ea02d91f313fb1a9c6aec.
> >> 
> >> Commits from @umn.edu addresses have been found to be submitted in "bad
> >> faith" to try to test the kernel community's ability to review "known
> >> malicious" changes.  The result of these submissions can be found in a
> >> paper published at the 42nd IEEE Symposium on Security and Privacy
> >> entitled, "Open Source Insecurity: Stealthily Introducing
> >> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> >> of Minnesota) and Kangjie Lu (University of Minnesota).
> >> 
> >> Because of this, all submissions from this group must be reverted from
> >> the kernel tree and will need to be re-reviewed again to determine if
> >> they actually are a valid fix.  Until that work is complete, remove this
> >> change to ensure that no problems are being introduced into the
> >> codebase.
> >> 
> >> Cc: stable@vger.kernel.org # v3.16+
> >> Cc: Guoqing Jiang <gqjiang@suse.com>
> >> Cc: Aditya Pakki <pakki001@umn.edu>
> >> Cc: Song Liu <songliubraving@fb.com>
> >> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> >> ---
> >> drivers/md/raid10.c | 2 --
> >> drivers/md/raid5.c  | 2 --
> >> 2 files changed, 4 deletions(-)
> >> 
> >> diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
> >> index a9ae7d113492..4fec1cdd4207 100644
> >> --- a/drivers/md/raid10.c
> >> +++ b/drivers/md/raid10.c
> >> @@ -3896,8 +3896,6 @@ static int raid10_run(struct mddev *mddev)
> >> 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
> >> 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
> >> 							"reshape");
> >> -		if (!mddev->sync_thread)
> >> -			goto out_free_conf;
> >> 	}
> >> 
> >> 	return 0;
> >> diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
> >> index 5d57a5bd171f..9b2bd50beee7 100644
> >> --- a/drivers/md/raid5.c
> >> +++ b/drivers/md/raid5.c
> >> @@ -7677,8 +7677,6 @@ static int raid5_run(struct mddev *mddev)
> >> 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
> >> 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
> >> 							"reshape");
> >> -		if (!mddev->sync_thread)
> >> -			goto abort;
> >> 	}
> >> 
> >> 	/* Ok, everything is just fine now */
> >> -- 
> >> 2.31.1
> >> 
> > 
> > These changes look ok, but the error handling logic seems to be freeing
> > the incorrect thread, not the one that these functions create.  That's
> > independant of this change, but seems odd.  If someone cares about it,
> > it should probably be looked at, or if correct, a comment would be nice
> > as it's really confusing.
> 
> I don't think this is confusing. raid[5|10]_run() creates two threads:
> first mddev->thread, then mddev->sync_thread. If we fail to create the
> second thread (sync_thread), we free the first thread (mddev->thread) in 
> the error handling logic. 

Look a bit lower, in "abort:" you do clean up "->thread", but never
"->sync_thread"  You can get to "abort:" after sync_thread is properly
registered.

That is what I was trying to say above, as it's not obvious because
"->thread" was created in a different function, setup_conf(), to see it
be freed here in the error path of raid[5|10]_run() wasn't all that
clear to me.

Anyway, these are code paths that are obviously never hit, so it's
probably not a big deal.

thanks,

greg k-h

[PATCH 181/190] Revert "dm ioctl: harden copy_params()'s copy_from_user() from malicious users"

[Greg Kroah-Hartman]

This reverts commit 800a7340ab7dd667edf95e74d8e4f23a17e87076.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: stable@vger.kernel.org
Cc: Wenwen Wang <wang6495@umn.edu>
Cc: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/md/dm-ioctl.c | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)

diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index 1ca65b434f1f..820342de92cd 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -1747,7 +1747,8 @@ static void free_params(struct dm_ioctl *param, size_t param_size, int param_fla
 }
 
 static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kernel,
-		       int ioctl_flags, struct dm_ioctl **param, int *param_flags)
+		       int ioctl_flags,
+		       struct dm_ioctl **param, int *param_flags)
 {
 	struct dm_ioctl *dmi;
 	int secure_data;
@@ -1788,13 +1789,18 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kern
 
 	*param_flags |= DM_PARAMS_MALLOC;
 
-	/* Copy from param_kernel (which was already copied from user) */
-	memcpy(dmi, param_kernel, minimum_data_size);
-
-	if (copy_from_user(&dmi->data, (char __user *)user + minimum_data_size,
-			   param_kernel->data_size - minimum_data_size))
+	if (copy_from_user(dmi, user, param_kernel->data_size))
 		goto bad;
+
 data_copied:
+	/*
+	 * Abort if something changed the ioctl data while it was being copied.
+	 */
+	if (dmi->data_size != param_kernel->data_size) {
+		DMERR("rejecting ioctl: data size modified while processing parameters");
+		goto bad;
+	}
+
 	/* Wipe the user buffer so we do not return it to userspace */
 	if (secure_data && clear_user(user, param_kernel->data_size))
 		goto bad;
-- 
2.31.1

Re: [PATCH 181/190] Revert "dm ioctl: harden copy_params()'s copy_from_user() from malicious users"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 03:00:56PM +0200, Greg Kroah-Hartman wrote:
> This reverts commit 800a7340ab7dd667edf95e74d8e4f23a17e87076.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: stable@vger.kernel.org
> Cc: Wenwen Wang <wang6495@umn.edu>
> Cc: Mike Snitzer <snitzer@redhat.com>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
>  drivers/md/dm-ioctl.c | 18 ++++++++++++------
>  1 file changed, 12 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
> index 1ca65b434f1f..820342de92cd 100644
> --- a/drivers/md/dm-ioctl.c
> +++ b/drivers/md/dm-ioctl.c
> @@ -1747,7 +1747,8 @@ static void free_params(struct dm_ioctl *param, size_t param_size, int param_fla
>  }
>  
>  static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kernel,
> -		       int ioctl_flags, struct dm_ioctl **param, int *param_flags)
> +		       int ioctl_flags,
> +		       struct dm_ioctl **param, int *param_flags)
>  {
>  	struct dm_ioctl *dmi;
>  	int secure_data;
> @@ -1788,13 +1789,18 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kern
>  
>  	*param_flags |= DM_PARAMS_MALLOC;
>  
> -	/* Copy from param_kernel (which was already copied from user) */
> -	memcpy(dmi, param_kernel, minimum_data_size);
> -
> -	if (copy_from_user(&dmi->data, (char __user *)user + minimum_data_size,
> -			   param_kernel->data_size - minimum_data_size))
> +	if (copy_from_user(dmi, user, param_kernel->data_size))
>  		goto bad;
> +
>  data_copied:
> +	/*
> +	 * Abort if something changed the ioctl data while it was being copied.
> +	 */
> +	if (dmi->data_size != param_kernel->data_size) {
> +		DMERR("rejecting ioctl: data size modified while processing parameters");
> +		goto bad;
> +	}
> +
>  	/* Wipe the user buffer so we do not return it to userspace */
>  	if (secure_data && clear_user(user, param_kernel->data_size))
>  		goto bad;
> -- 
> 2.31.1
> 

Original looks correct, dropping this commit now.

greg k-h

[PATCH 187/190] Revert "ALSA: control: fix a redundant-copy issue"

[Greg Kroah-Hartman]

This reverts commit 3f12888dfae2a48741c4caa9214885b3aaf350f9.

Commits from @umn.edu addresses have been found to be submitted in "bad
faith" to try to test the kernel community's ability to review "known
malicious" changes.  The result of these submissions can be found in a
paper published at the 42nd IEEE Symposium on Security and Privacy
entitled, "Open Source Insecurity: Stealthily Introducing
Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
of Minnesota) and Kangjie Lu (University of Minnesota).

Because of this, all submissions from this group must be reverted from
the kernel tree and will need to be re-reviewed again to determine if
they actually are a valid fix.  Until that work is complete, remove this
change to ensure that no problems are being introduced into the
codebase.

Cc: Wenwen Wang <wang6495@umn.edu>
Cc: <stable@vger.kernel.org>
Cc: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 sound/core/control_compat.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/sound/core/control_compat.c b/sound/core/control_compat.c
index 1d708aab9c98..857acf83ae47 100644
--- a/sound/core/control_compat.c
+++ b/sound/core/control_compat.c
@@ -381,7 +381,8 @@ static int snd_ctl_elem_add_compat(struct snd_ctl_file *file,
 	if (copy_from_user(&data->id, &data32->id, sizeof(data->id)) ||
 	    copy_from_user(&data->type, &data32->type, 3 * sizeof(u32)))
 		goto error;
-	if (get_user(data->owner, &data32->owner))
+	if (get_user(data->owner, &data32->owner) ||
+	    get_user(data->type, &data32->type))
 		goto error;
 	switch (data->type) {
 	case SNDRV_CTL_ELEM_TYPE_BOOLEAN:
-- 
2.31.1

Re: [PATCH 187/190] Revert "ALSA: control: fix a redundant-copy issue"

[Takashi Iwai]

On Wed, 21 Apr 2021 15:01:02 +0200,
Greg Kroah-Hartman wrote:
> 
> This reverts commit 3f12888dfae2a48741c4caa9214885b3aaf350f9.
> 
> Commits from @umn.edu addresses have been found to be submitted in "bad
> faith" to try to test the kernel community's ability to review "known
> malicious" changes.  The result of these submissions can be found in a
> paper published at the 42nd IEEE Symposium on Security and Privacy
> entitled, "Open Source Insecurity: Stealthily Introducing
> Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> of Minnesota) and Kangjie Lu (University of Minnesota).
> 
> Because of this, all submissions from this group must be reverted from
> the kernel tree and will need to be re-reviewed again to determine if
> they actually are a valid fix.  Until that work is complete, remove this
> change to ensure that no problems are being introduced into the
> codebase.
> 
> Cc: Wenwen Wang <wang6495@umn.edu>
> Cc: <stable@vger.kernel.org>
> Cc: Takashi Iwai <tiwai@suse.de>
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This one is, unlike other patches I've been involved with, about the
ALSA core code, and this change is likely worth to keep.

The code change is correct, and even though it's really a minor issue,
an optimization is right.


thanks,

Takashi


> ---
>  sound/core/control_compat.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/sound/core/control_compat.c b/sound/core/control_compat.c
> index 1d708aab9c98..857acf83ae47 100644
> --- a/sound/core/control_compat.c
> +++ b/sound/core/control_compat.c
> @@ -381,7 +381,8 @@ static int snd_ctl_elem_add_compat(struct snd_ctl_file *file,
>  	if (copy_from_user(&data->id, &data32->id, sizeof(data->id)) ||
>  	    copy_from_user(&data->type, &data32->type, 3 * sizeof(u32)))
>  		goto error;
> -	if (get_user(data->owner, &data32->owner))
> +	if (get_user(data->owner, &data32->owner) ||
> +	    get_user(data->type, &data32->type))
>  		goto error;
>  	switch (data->type) {
>  	case SNDRV_CTL_ELEM_TYPE_BOOLEAN:
> -- 
> 2.31.1
> 

Re: [PATCH 187/190] Revert "ALSA: control: fix a redundant-copy issue"

[Greg Kroah-Hartman]

On Wed, Apr 21, 2021 at 06:30:07PM +0200, Takashi Iwai wrote:
> On Wed, 21 Apr 2021 15:01:02 +0200,
> Greg Kroah-Hartman wrote:
> > 
> > This reverts commit 3f12888dfae2a48741c4caa9214885b3aaf350f9.
> > 
> > Commits from @umn.edu addresses have been found to be submitted in "bad
> > faith" to try to test the kernel community's ability to review "known
> > malicious" changes.  The result of these submissions can be found in a
> > paper published at the 42nd IEEE Symposium on Security and Privacy
> > entitled, "Open Source Insecurity: Stealthily Introducing
> > Vulnerabilities via Hypocrite Commits" written by Qiushi Wu (University
> > of Minnesota) and Kangjie Lu (University of Minnesota).
> > 
> > Because of this, all submissions from this group must be reverted from
> > the kernel tree and will need to be re-reviewed again to determine if
> > they actually are a valid fix.  Until that work is complete, remove this
> > change to ensure that no problems are being introduced into the
> > codebase.
> > 
> > Cc: Wenwen Wang <wang6495@umn.edu>
> > Cc: <stable@vger.kernel.org>
> > Cc: Takashi Iwai <tiwai@suse.de>
> > Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> 
> This one is, unlike other patches I've been involved with, about the
> ALSA core code, and this change is likely worth to keep.
> 
> The code change is correct, and even though it's really a minor issue,
> an optimization is right.

Thanks for the review, now dropped.

greg k-h