[U-Boot] [PATCH v2 0/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL to whole disk

[U-Boot] [PATCH v2 0/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL to whole disk

[AKASHI Takahiro]

Changes in v2 (Sept 12, 2019)
* add patch#1 and #2
* install the protocol only if a file system does exist

AKASHI Takahiro (3):
  fs: export fs_close()
  efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available
  efi_loader: disk: install file system protocol to a whole disk

 fs/fs.c                   |  2 +-
 include/fs.h              |  7 +++++++
 lib/efi_loader/efi_disk.c | 19 ++++++++++++++++++-
 3 files changed, 26 insertions(+), 2 deletions(-)

-- 
2.21.0

[U-Boot] [PATCH v2 1/3] fs: export fs_close()

[AKASHI Takahiro]

This function is always paired with either fs_set_blk_desc() or
fs_set_blk_desc_with_part(). So just export it.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
---
 fs/fs.c      | 2 +-
 include/fs.h | 7 +++++++
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/fs/fs.c b/fs/fs.c
index d8a4ced4698e..64ba25fea8bf 100644
--- a/fs/fs.c
+++ b/fs/fs.c
@@ -389,7 +389,7 @@ int fs_set_blk_dev_with_part(struct blk_desc *desc, int part)
 	return -1;
 }
 
-static void fs_close(void)
+void fs_close(void)
 {
 	struct fstype_info *info = fs_get_info(fs_type);
 
diff --git a/include/fs.h b/include/fs.h
index 7601b0343bcd..5a1244d57fd2 100644
--- a/include/fs.h
+++ b/include/fs.h
@@ -37,6 +37,13 @@ int fs_set_blk_dev(const char *ifname, const char *dev_part_str, int fstype);
  */
 int fs_set_blk_dev_with_part(struct blk_desc *desc, int part);
 
+/**
+ * fs_close() - Unset current block device and partition
+ *
+ * Should be paired with either fs_set_blk_dev() or fs_set_dev_with_part()
+ */
+void fs_close(void);
+
 /**
  * fs_get_type_name() - Get type of current filesystem
  *
-- 
2.21.0

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[AKASHI Takahiro]

In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
installed to all the partitions even if some of them may house no file
system.

With this patch, that protocol will be installed only if FAT file system
exists.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
---
 lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
index 7a6b06821a47..d72f455901f2 100644
--- a/lib/efi_loader/efi_disk.c
+++ b/lib/efi_loader/efi_disk.c
@@ -9,6 +9,7 @@
 #include <blk.h>
 #include <dm.h>
 #include <efi_loader.h>
+#include <fs.h>
 #include <part.h>
 #include <malloc.h>
 
@@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
 	return handler->protocol_interface;
 }
 
+static int efi_fs_exists(struct blk_desc *desc, int part)
+{
+	if (fs_set_blk_dev_with_part(desc, part))
+		return 0;
+
+	if (strcmp(fs_get_type_name(), "fat"))
+		return 0;
+
+	fs_close();
+
+	return 1;
+}
+
 /*
  * Create a handle for a partition or disk
  *
@@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
 			       diskobj->dp);
 	if (ret != EFI_SUCCESS)
 		return ret;
-	if (part >= 1) {
+	if (part >= 1 && efi_fs_exists(desc, part)) {
 		diskobj->volume = efi_simple_file_system(desc, part,
 							 diskobj->dp);
 		ret = efi_add_protocol(&diskobj->header,
-- 
2.21.0

[U-Boot] [PATCH v2 3/3] efi_loader: disk: install file system protocol to a whole disk

[AKASHI Takahiro]

Currently, a whole disk without any partitions is not associated
with EFI_SIMPLE_FILE_SYSTEM_PROTOCOL. So even if it houses FAT
file system, there is a chance that we may not be able to access
it, particularly, when accesses are to be attempted after searching
that protocol against a device handle.

With this patch, EFI_SIMPLE_FILE_SYSTEM_PROTOCOL is installed
to such a disk if part_get_info() shows there is not partition
table installed on it.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
---
 lib/efi_loader/efi_disk.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
index d72f455901f2..d36f22cedc52 100644
--- a/lib/efi_loader/efi_disk.c
+++ b/lib/efi_loader/efi_disk.c
@@ -253,6 +253,7 @@ static efi_status_t efi_disk_add_dev(
 				struct efi_disk_obj **disk)
 {
 	struct efi_disk_obj *diskobj;
+	disk_partition_t info;
 	efi_status_t ret;
 
 	/* Don't add empty devices */
@@ -284,7 +285,9 @@ static efi_status_t efi_disk_add_dev(
 			       diskobj->dp);
 	if (ret != EFI_SUCCESS)
 		return ret;
-	if (part >= 1 && efi_fs_exists(desc, part)) {
+	/* partitions or whole disk without partitions */
+	if ((part >= 1 || part_get_info(desc, part, &info)) &&
+	    efi_fs_exists(desc, part)) {
 		diskobj->volume = efi_simple_file_system(desc, part,
 							 diskobj->dp);
 		ret = efi_add_protocol(&diskobj->header,
-- 
2.21.0

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[Heinrich Schuchardt]

On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
> installed to all the partitions even if some of them may house no file
> system.
>
> With this patch, that protocol will be installed only if FAT file system
> exists.
>
> Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> ---
>  lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
>  1 file changed, 15 insertions(+), 1 deletion(-)
>
> diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> index 7a6b06821a47..d72f455901f2 100644
> --- a/lib/efi_loader/efi_disk.c
> +++ b/lib/efi_loader/efi_disk.c
> @@ -9,6 +9,7 @@
>  #include <blk.h>
>  #include <dm.h>
>  #include <efi_loader.h>
> +#include <fs.h>
>  #include <part.h>
>  #include <malloc.h>
>
> @@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
>  	return handler->protocol_interface;
>  }
>
> +static int efi_fs_exists(struct blk_desc *desc, int part)
> +{
> +	if (fs_set_blk_dev_with_part(desc, part))
> +		return 0;
> +
> +	if (strcmp(fs_get_type_name(), "fat"))

Before your patch we could use any supported file system (e.g. EXT2). I
see no need for a restriction to FAT. You could compare the string to
"unsupported":

if (!strcmp(fs_get_type_name(), "unsupported"))
	return 0;

But wouldn't it be preferable to have a function to access fs_type (in
fs/fs.c) directly instead of a string representation?

Otherwise we should convert the string "unsupported" of fstypes[] into a
constant in fs.h so that we can be sure we are using the same value.

Best regards

Heinrich

> +		return 0;
> +
> +	fs_close();
> +
> +	return 1;
> +}
> +
>  /*
>   * Create a handle for a partition or disk
>   *
> @@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
>  			       diskobj->dp);
>  	if (ret != EFI_SUCCESS)
>  		return ret;
> -	if (part >= 1) {
> +	if (part >= 1 && efi_fs_exists(desc, part)) {
>  		diskobj->volume = efi_simple_file_system(desc, part,
>  							 diskobj->dp);
>  		ret = efi_add_protocol(&diskobj->header,
>

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[AKASHI Takahiro]

On Thu, Sep 12, 2019 at 10:57:20AM +0200, Heinrich Schuchardt wrote:
> On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> > In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
> > installed to all the partitions even if some of them may house no file
> > system.
> >
> > With this patch, that protocol will be installed only if FAT file system
> > exists.
> >
> > Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> > ---
> >  lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
> >  1 file changed, 15 insertions(+), 1 deletion(-)
> >
> > diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> > index 7a6b06821a47..d72f455901f2 100644
> > --- a/lib/efi_loader/efi_disk.c
> > +++ b/lib/efi_loader/efi_disk.c
> > @@ -9,6 +9,7 @@
> >  #include <blk.h>
> >  #include <dm.h>
> >  #include <efi_loader.h>
> > +#include <fs.h>
> >  #include <part.h>
> >  #include <malloc.h>
> >
> > @@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
> >  	return handler->protocol_interface;
> >  }
> >
> > +static int efi_fs_exists(struct blk_desc *desc, int part)
> > +{
> > +	if (fs_set_blk_dev_with_part(desc, part))
> > +		return 0;
> > +
> > +	if (strcmp(fs_get_type_name(), "fat"))
> 
> Before your patch we could use any supported file system (e.g. EXT2). I
> see no need for a restriction to FAT. You could compare the string to
> "unsupported":

No. As far as you want to stick to compliance to UEFI specification,
"fat" is the only file system supported by UEFI.

> if (!strcmp(fs_get_type_name(), "unsupported"))
> 	return 0;
> 
> But wouldn't it be preferable to have a function to access fs_type (in
> fs/fs.c) directly instead of a string representation?

Agree, but there is no direct function in fs/fs.c.
I'm reluctant to invent a new function just for this purpose.

-Takahiro Akashi


> Otherwise we should convert the string "unsupported" of fstypes[] into a
> constant in fs.h so that we can be sure we are using the same value.
> 
> Best regards
> 
> Heinrich
> 
> > +		return 0;
> > +
> > +	fs_close();
> > +
> > +	return 1;
> > +}
> > +
> >  /*
> >   * Create a handle for a partition or disk
> >   *
> > @@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
> >  			       diskobj->dp);
> >  	if (ret != EFI_SUCCESS)
> >  		return ret;
> > -	if (part >= 1) {
> > +	if (part >= 1 && efi_fs_exists(desc, part)) {
> >  		diskobj->volume = efi_simple_file_system(desc, part,
> >  							 diskobj->dp);
> >  		ret = efi_add_protocol(&diskobj->header,
> >
> 

[U-Boot] [PATCH v2 3/3] efi_loader: disk: install file system protocol to a whole disk

[Heinrich Schuchardt]

On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> Currently, a whole disk without any partitions is not associated
> with EFI_SIMPLE_FILE_SYSTEM_PROTOCOL. So even if it houses FAT
> file system, there is a chance that we may not be able to access
> it, particularly, when accesses are to be attempted after searching
> that protocol against a device handle.
>
> With this patch, EFI_SIMPLE_FILE_SYSTEM_PROTOCOL is installed
> to such a disk if part_get_info() shows there is not partition
> table installed on it.
>
> Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> ---
>  lib/efi_loader/efi_disk.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> index d72f455901f2..d36f22cedc52 100644
> --- a/lib/efi_loader/efi_disk.c
> +++ b/lib/efi_loader/efi_disk.c
> @@ -253,6 +253,7 @@ static efi_status_t efi_disk_add_dev(
>  				struct efi_disk_obj **disk)
>  {
>  	struct efi_disk_obj *diskobj;
> +	disk_partition_t info;
>  	efi_status_t ret;
>
>  	/* Don't add empty devices */
> @@ -284,7 +285,9 @@ static efi_status_t efi_disk_add_dev(
>  			       diskobj->dp);
>  	if (ret != EFI_SUCCESS)
>  		return ret;
> -	if (part >= 1 && efi_fs_exists(desc, part)) {
> +	/* partitions or whole disk without partitions */
> +	if ((part >= 1 || part_get_info(desc, part, &info)) &&

part_get_info() returns -1 for part = 0 on a DOS partioned disk. So this
check does not work.

Best regards

Heinrich

> +	    efi_fs_exists(desc, part)) {
>  		diskobj->volume = efi_simple_file_system(desc, part,
>  							 diskobj->dp);
>  		ret = efi_add_protocol(&diskobj->header,
>

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[Heinrich Schuchardt]

On 9/12/19 11:17 AM, AKASHI Takahiro wrote:
> On Thu, Sep 12, 2019 at 10:57:20AM +0200, Heinrich Schuchardt wrote:
>> On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
>>> In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
>>> installed to all the partitions even if some of them may house no file
>>> system.
>>>
>>> With this patch, that protocol will be installed only if FAT file system
>>> exists.
>>>
>>> Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
>>> ---
>>>  lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
>>>  1 file changed, 15 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
>>> index 7a6b06821a47..d72f455901f2 100644
>>> --- a/lib/efi_loader/efi_disk.c
>>> +++ b/lib/efi_loader/efi_disk.c
>>> @@ -9,6 +9,7 @@
>>>  #include <blk.h>
>>>  #include <dm.h>
>>>  #include <efi_loader.h>
>>> +#include <fs.h>
>>>  #include <part.h>
>>>  #include <malloc.h>
>>>
>>> @@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
>>>  	return handler->protocol_interface;
>>>  }
>>>
>>> +static int efi_fs_exists(struct blk_desc *desc, int part)
>>> +{
>>> +	if (fs_set_blk_dev_with_part(desc, part))
>>> +		return 0;
>>> +
>>> +	if (strcmp(fs_get_type_name(), "fat"))
>>
>> Before your patch we could use any supported file system (e.g. EXT2). I
>> see no need for a restriction to FAT. You could compare the string to
>> "unsupported":
>
> No. As far as you want to stick to compliance to UEFI specification,
> "fat" is the only file system supported by UEFI.

In the case of device path node VenHw() there is a direct rule in the
spec indicating how it should be rendered. I have not seen anything in
the UEFI spec saying that you should not support file systems besides
FAT. So there is no compliance issue. I would be reluctant to remove an
existing capability of U-Boot.

>
>> if (!strcmp(fs_get_type_name(), "unsupported"))
>> 	return 0;
>>
>> But wouldn't it be preferable to have a function to access fs_type (in
>> fs/fs.c) directly instead of a string representation?
>
> Agree, but there is no direct function in fs/fs.c.
> I'm reluctant to invent a new function just for this purpose.

In that case we should compare to a string that is defined as constant
in fs.h.

Best regards

Heinrich

>
> -Takahiro Akashi
>
>
>> Otherwise we should convert the string "unsupported" of fstypes[] into a
>> constant in fs.h so that we can be sure we are using the same value.
>>
>> Best regards
>>
>> Heinrich
>>
>>> +		return 0;
>>> +
>>> +	fs_close();
>>> +
>>> +	return 1;
>>> +}
>>> +
>>>  /*
>>>   * Create a handle for a partition or disk
>>>   *
>>> @@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
>>>  			       diskobj->dp);
>>>  	if (ret != EFI_SUCCESS)
>>>  		return ret;
>>> -	if (part >= 1) {
>>> +	if (part >= 1 && efi_fs_exists(desc, part)) {
>>>  		diskobj->volume = efi_simple_file_system(desc, part,
>>>  							 diskobj->dp);
>>>  		ret = efi_add_protocol(&diskobj->header,
>>>
>>
>

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[AKASHI Takahiro]

On Thu, Sep 12, 2019 at 11:43:07AM +0200, Heinrich Schuchardt wrote:
> On 9/12/19 11:17 AM, AKASHI Takahiro wrote:
> > On Thu, Sep 12, 2019 at 10:57:20AM +0200, Heinrich Schuchardt wrote:
> >> On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> >>> In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
> >>> installed to all the partitions even if some of them may house no file
> >>> system.
> >>>
> >>> With this patch, that protocol will be installed only if FAT file system
> >>> exists.
> >>>
> >>> Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> >>> ---
> >>>  lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
> >>>  1 file changed, 15 insertions(+), 1 deletion(-)
> >>>
> >>> diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> >>> index 7a6b06821a47..d72f455901f2 100644
> >>> --- a/lib/efi_loader/efi_disk.c
> >>> +++ b/lib/efi_loader/efi_disk.c
> >>> @@ -9,6 +9,7 @@
> >>>  #include <blk.h>
> >>>  #include <dm.h>
> >>>  #include <efi_loader.h>
> >>> +#include <fs.h>
> >>>  #include <part.h>
> >>>  #include <malloc.h>
> >>>
> >>> @@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
> >>>  	return handler->protocol_interface;
> >>>  }
> >>>
> >>> +static int efi_fs_exists(struct blk_desc *desc, int part)
> >>> +{
> >>> +	if (fs_set_blk_dev_with_part(desc, part))
> >>> +		return 0;
> >>> +
> >>> +	if (strcmp(fs_get_type_name(), "fat"))
> >>
> >> Before your patch we could use any supported file system (e.g. EXT2). I
> >> see no need for a restriction to FAT. You could compare the string to
> >> "unsupported":
> >
> > No. As far as you want to stick to compliance to UEFI specification,
> > "fat" is the only file system supported by UEFI.
> 
> In the case of device path node VenHw() there is a direct rule in the
> spec indicating how it should be rendered. I have not seen anything in
> the UEFI spec saying that you should not support file systems besides
> FAT. So there is no compliance issue. I would be reluctant to remove an
> existing capability of U-Boot.

See section 13.3. It says,
        The file system supported by the Extensible Firmware Interface is
        based on the FAT file system.

-Takahiro Akashi

> >
> >> if (!strcmp(fs_get_type_name(), "unsupported"))
> >> 	return 0;
> >>
> >> But wouldn't it be preferable to have a function to access fs_type (in
> >> fs/fs.c) directly instead of a string representation?
> >
> > Agree, but there is no direct function in fs/fs.c.
> > I'm reluctant to invent a new function just for this purpose.
> 
> In that case we should compare to a string that is defined as constant
> in fs.h.
> 
> Best regards
> 
> Heinrich
> 
> >
> > -Takahiro Akashi
> >
> >
> >> Otherwise we should convert the string "unsupported" of fstypes[] into a
> >> constant in fs.h so that we can be sure we are using the same value.
> >>
> >> Best regards
> >>
> >> Heinrich
> >>
> >>> +		return 0;
> >>> +
> >>> +	fs_close();
> >>> +
> >>> +	return 1;
> >>> +}
> >>> +
> >>>  /*
> >>>   * Create a handle for a partition or disk
> >>>   *
> >>> @@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
> >>>  			       diskobj->dp);
> >>>  	if (ret != EFI_SUCCESS)
> >>>  		return ret;
> >>> -	if (part >= 1) {
> >>> +	if (part >= 1 && efi_fs_exists(desc, part)) {
> >>>  		diskobj->volume = efi_simple_file_system(desc, part,
> >>>  							 diskobj->dp);
> >>>  		ret = efi_add_protocol(&diskobj->header,
> >>>
> >>
> >
> 

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[AKASHI Takahiro]

Heinrich,

On Thu, Sep 12, 2019 at 06:51:35PM +0900, AKASHI Takahiro wrote:
> On Thu, Sep 12, 2019 at 11:43:07AM +0200, Heinrich Schuchardt wrote:
> > On 9/12/19 11:17 AM, AKASHI Takahiro wrote:
> > > On Thu, Sep 12, 2019 at 10:57:20AM +0200, Heinrich Schuchardt wrote:
> > >> On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> > >>> In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
> > >>> installed to all the partitions even if some of them may house no file
> > >>> system.
> > >>>
> > >>> With this patch, that protocol will be installed only if FAT file system
> > >>> exists.
> > >>>
> > >>> Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> > >>> ---
> > >>>  lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
> > >>>  1 file changed, 15 insertions(+), 1 deletion(-)
> > >>>
> > >>> diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> > >>> index 7a6b06821a47..d72f455901f2 100644
> > >>> --- a/lib/efi_loader/efi_disk.c
> > >>> +++ b/lib/efi_loader/efi_disk.c
> > >>> @@ -9,6 +9,7 @@
> > >>>  #include <blk.h>
> > >>>  #include <dm.h>
> > >>>  #include <efi_loader.h>
> > >>> +#include <fs.h>
> > >>>  #include <part.h>
> > >>>  #include <malloc.h>
> > >>>
> > >>> @@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
> > >>>  	return handler->protocol_interface;
> > >>>  }
> > >>>
> > >>> +static int efi_fs_exists(struct blk_desc *desc, int part)
> > >>> +{
> > >>> +	if (fs_set_blk_dev_with_part(desc, part))
> > >>> +		return 0;
> > >>> +
> > >>> +	if (strcmp(fs_get_type_name(), "fat"))
> > >>
> > >> Before your patch we could use any supported file system (e.g. EXT2). I
> > >> see no need for a restriction to FAT. You could compare the string to
> > >> "unsupported":
> > >
> > > No. As far as you want to stick to compliance to UEFI specification,
> > > "fat" is the only file system supported by UEFI.
> > 
> > In the case of device path node VenHw() there is a direct rule in the
> > spec indicating how it should be rendered. I have not seen anything in
> > the UEFI spec saying that you should not support file systems besides
> > FAT. So there is no compliance issue. I would be reluctant to remove an
> > existing capability of U-Boot.
> 
> See section 13.3. It says,
>         The file system supported by the Extensible Firmware Interface is
>         based on the FAT file system.

Any comments here?

> > >
> > >> if (!strcmp(fs_get_type_name(), "unsupported"))
> > >> 	return 0;
> > >>
> > >> But wouldn't it be preferable to have a function to access fs_type (in
> > >> fs/fs.c) directly instead of a string representation?
> > >
> > > Agree, but there is no direct function in fs/fs.c.
> > > I'm reluctant to invent a new function just for this purpose.
> > 
> > In that case we should compare to a string that is defined as constant
> > in fs.h.

Are you saying that we should add new macros in fs.h?
All the file system names are initialized in fs/fs.c now.

> -Takahiro Akashi
> 
> > 
> > Best regards
> > 
> > Heinrich
> > 
> > >
> > > -Takahiro Akashi
> > >
> > >
> > >> Otherwise we should convert the string "unsupported" of fstypes[] into a
> > >> constant in fs.h so that we can be sure we are using the same value.
> > >>
> > >> Best regards
> > >>
> > >> Heinrich
> > >>
> > >>> +		return 0;
> > >>> +
> > >>> +	fs_close();
> > >>> +
> > >>> +	return 1;
> > >>> +}
> > >>> +
> > >>>  /*
> > >>>   * Create a handle for a partition or disk
> > >>>   *
> > >>> @@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
> > >>>  			       diskobj->dp);
> > >>>  	if (ret != EFI_SUCCESS)
> > >>>  		return ret;
> > >>> -	if (part >= 1) {
> > >>> +	if (part >= 1 && efi_fs_exists(desc, part)) {
> > >>>  		diskobj->volume = efi_simple_file_system(desc, part,
> > >>>  							 diskobj->dp);
> > >>>  		ret = efi_add_protocol(&diskobj->header,
> > >>>
> > >>
> > >
> > 

[U-Boot] [PATCH v2 3/3] efi_loader: disk: install file system protocol to a whole disk

[AKASHI Takahiro]

Heinrich,

On Thu, Sep 12, 2019 at 11:21:51AM +0200, Heinrich Schuchardt wrote:
> On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> > Currently, a whole disk without any partitions is not associated
> > with EFI_SIMPLE_FILE_SYSTEM_PROTOCOL. So even if it houses FAT
> > file system, there is a chance that we may not be able to access
> > it, particularly, when accesses are to be attempted after searching
> > that protocol against a device handle.
> >
> > With this patch, EFI_SIMPLE_FILE_SYSTEM_PROTOCOL is installed
> > to such a disk if part_get_info() shows there is not partition
> > table installed on it.
> >
> > Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> > ---
> >  lib/efi_loader/efi_disk.c | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> > index d72f455901f2..d36f22cedc52 100644
> > --- a/lib/efi_loader/efi_disk.c
> > +++ b/lib/efi_loader/efi_disk.c
> > @@ -253,6 +253,7 @@ static efi_status_t efi_disk_add_dev(
> >  				struct efi_disk_obj **disk)
> >  {
> >  	struct efi_disk_obj *diskobj;
> > +	disk_partition_t info;
> >  	efi_status_t ret;
> >
> >  	/* Don't add empty devices */
> > @@ -284,7 +285,9 @@ static efi_status_t efi_disk_add_dev(
> >  			       diskobj->dp);
> >  	if (ret != EFI_SUCCESS)
> >  		return ret;
> > -	if (part >= 1 && efi_fs_exists(desc, part)) {
> > +	/* partitions or whole disk without partitions */
> > +	if ((part >= 1 || part_get_info(desc, part, &info)) &&
> 
> part_get_info() returns -1 for part = 0 on a DOS partioned disk. So this
> check does not work.

Right.
So this statement should be simplified like:
        if (efi_fs_exists(desc, part)) {
                diskobj->volume = efi_simple_file_system(desc, part,
                                                         diskobj->dp);
                ret = efi_add_protocol(&diskobj->header,
                ...
        }

-Takahiro Akashi


> Best regards
> 
> Heinrich
> 
> > +	    efi_fs_exists(desc, part)) {
> >  		diskobj->volume = efi_simple_file_system(desc, part,
> >  							 diskobj->dp);
> >  		ret = efi_add_protocol(&diskobj->header,
> >
> 

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[Heinrich Schuchardt]

On 10/3/19 9:09 AM, AKASHI Takahiro wrote:
> Heinrich,
>
> On Thu, Sep 12, 2019 at 06:51:35PM +0900, AKASHI Takahiro wrote:
>> On Thu, Sep 12, 2019 at 11:43:07AM +0200, Heinrich Schuchardt wrote:
>>> On 9/12/19 11:17 AM, AKASHI Takahiro wrote:
>>>> On Thu, Sep 12, 2019 at 10:57:20AM +0200, Heinrich Schuchardt wrote:
>>>>> On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
>>>>>> In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
>>>>>> installed to all the partitions even if some of them may house no file
>>>>>> system.
>>>>>>
>>>>>> With this patch, that protocol will be installed only if FAT file system
>>>>>> exists.
>>>>>>
>>>>>> Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
>>>>>> ---
>>>>>>   lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
>>>>>>   1 file changed, 15 insertions(+), 1 deletion(-)
>>>>>>
>>>>>> diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
>>>>>> index 7a6b06821a47..d72f455901f2 100644
>>>>>> --- a/lib/efi_loader/efi_disk.c
>>>>>> +++ b/lib/efi_loader/efi_disk.c
>>>>>> @@ -9,6 +9,7 @@
>>>>>>   #include <blk.h>
>>>>>>   #include <dm.h>
>>>>>>   #include <efi_loader.h>
>>>>>> +#include <fs.h>
>>>>>>   #include <part.h>
>>>>>>   #include <malloc.h>
>>>>>>
>>>>>> @@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
>>>>>>   	return handler->protocol_interface;
>>>>>>   }
>>>>>>
>>>>>> +static int efi_fs_exists(struct blk_desc *desc, int part)
>>>>>> +{
>>>>>> +	if (fs_set_blk_dev_with_part(desc, part))
>>>>>> +		return 0;
>>>>>> +
>>>>>> +	if (strcmp(fs_get_type_name(), "fat"))
>>>>>
>>>>> Before your patch we could use any supported file system (e.g. EXT2). I
>>>>> see no need for a restriction to FAT. You could compare the string to
>>>>> "unsupported":
>>>>
>>>> No. As far as you want to stick to compliance to UEFI specification,
>>>> "fat" is the only file system supported by UEFI.
>>>
>>> In the case of device path node VenHw() there is a direct rule in the
>>> spec indicating how it should be rendered. I have not seen anything in
>>> the UEFI spec saying that you should not support file systems besides
>>> FAT. So there is no compliance issue. I would be reluctant to remove an
>>> existing capability of U-Boot.
>>
>> See section 13.3. It says,
>>          The file system supported by the Extensible Firmware Interface is
>>          based on the FAT file system.
>
> Any comments here?

The spec does not say: Thou shalt not support other file systems.

>
>>>>
>>>>> if (!strcmp(fs_get_type_name(), "unsupported"))
>>>>> 	return 0;
>>>>>
>>>>> But wouldn't it be preferable to have a function to access fs_type (in
>>>>> fs/fs.c) directly instead of a string representation?
>>>>
>>>> Agree, but there is no direct function in fs/fs.c.
>>>> I'm reluctant to invent a new function just for this purpose.
>>>
>>> In that case we should compare to a string that is defined as constant
>>> in fs.h.
>
> Are you saying that we should add new macros in fs.h?
> All the file system names are initialized in fs/fs.c now.
>
>> -Takahiro Akashi
>>
>>>
>>> Best regards
>>>
>>> Heinrich
>>>
>>>>
>>>> -Takahiro Akashi
>>>>
>>>>
>>>>> Otherwise we should convert the string "unsupported" of fstypes[] into a
>>>>> constant in fs.h so that we can be sure we are using the same value.
>>>>>
>>>>> Best regards
>>>>>
>>>>> Heinrich
>>>>>
>>>>>> +		return 0;
>>>>>> +
>>>>>> +	fs_close();
>>>>>> +
>>>>>> +	return 1;
>>>>>> +}
>>>>>> +
>>>>>>   /*
>>>>>>    * Create a handle for a partition or disk
>>>>>>    *
>>>>>> @@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
>>>>>>   			       diskobj->dp);
>>>>>>   	if (ret != EFI_SUCCESS)
>>>>>>   		return ret;
>>>>>> -	if (part >= 1) {
>>>>>> +	if (part >= 1 && efi_fs_exists(desc, part)) {
>>>>>>   		diskobj->volume = efi_simple_file_system(desc, part,
>>>>>>   							 diskobj->dp);
>>>>>>   		ret = efi_add_protocol(&diskobj->header,
>>>>>>
>>>>>
>>>>
>>>
>

[U-Boot] [PATCH v2 2/3] efi_loader: disk: install FILE_SYSTEM_PROTOCOL only if available

[AKASHI Takahiro]

On Thu, Oct 03, 2019 at 03:44:54PM +0200, Heinrich Schuchardt wrote:
> On 10/3/19 9:09 AM, AKASHI Takahiro wrote:
> >Heinrich,
> >
> >On Thu, Sep 12, 2019 at 06:51:35PM +0900, AKASHI Takahiro wrote:
> >>On Thu, Sep 12, 2019 at 11:43:07AM +0200, Heinrich Schuchardt wrote:
> >>>On 9/12/19 11:17 AM, AKASHI Takahiro wrote:
> >>>>On Thu, Sep 12, 2019 at 10:57:20AM +0200, Heinrich Schuchardt wrote:
> >>>>>On 9/12/19 6:51 AM, AKASHI Takahiro wrote:
> >>>>>>In the current implementation, EFI_SIMPLEFILE_SYSTEM_PROTOCOL is always
> >>>>>>installed to all the partitions even if some of them may house no file
> >>>>>>system.
> >>>>>>
> >>>>>>With this patch, that protocol will be installed only if FAT file system
> >>>>>>exists.
> >>>>>>
> >>>>>>Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
> >>>>>>---
> >>>>>>  lib/efi_loader/efi_disk.c | 16 +++++++++++++++-
> >>>>>>  1 file changed, 15 insertions(+), 1 deletion(-)
> >>>>>>
> >>>>>>diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c
> >>>>>>index 7a6b06821a47..d72f455901f2 100644
> >>>>>>--- a/lib/efi_loader/efi_disk.c
> >>>>>>+++ b/lib/efi_loader/efi_disk.c
> >>>>>>@@ -9,6 +9,7 @@
> >>>>>>  #include <blk.h>
> >>>>>>  #include <dm.h>
> >>>>>>  #include <efi_loader.h>
> >>>>>>+#include <fs.h>
> >>>>>>  #include <part.h>
> >>>>>>  #include <malloc.h>
> >>>>>>
> >>>>>>@@ -217,6 +218,19 @@ efi_fs_from_path(struct efi_device_path *full_path)
> >>>>>>  	return handler->protocol_interface;
> >>>>>>  }
> >>>>>>
> >>>>>>+static int efi_fs_exists(struct blk_desc *desc, int part)
> >>>>>>+{
> >>>>>>+	if (fs_set_blk_dev_with_part(desc, part))
> >>>>>>+		return 0;
> >>>>>>+
> >>>>>>+	if (strcmp(fs_get_type_name(), "fat"))
> >>>>>
> >>>>>Before your patch we could use any supported file system (e.g. EXT2). I
> >>>>>see no need for a restriction to FAT. You could compare the string to
> >>>>>"unsupported":
> >>>>
> >>>>No. As far as you want to stick to compliance to UEFI specification,
> >>>>"fat" is the only file system supported by UEFI.
> >>>
> >>>In the case of device path node VenHw() there is a direct rule in the
> >>>spec indicating how it should be rendered. I have not seen anything in
> >>>the UEFI spec saying that you should not support file systems besides
> >>>FAT. So there is no compliance issue. I would be reluctant to remove an
> >>>existing capability of U-Boot.
> >>
> >>See section 13.3. It says,
> >>         The file system supported by the Extensible Firmware Interface is
> >>         based on the FAT file system.
> >
> >Any comments here?
> 
> The spec does not say: Thou shalt not support other file systems.

If we admit such an excuse for arbitrary extension,
it will easily break compatibility in general.

-Takahiro Akashi


> >
> >>>>
> >>>>>if (!strcmp(fs_get_type_name(), "unsupported"))
> >>>>>	return 0;
> >>>>>
> >>>>>But wouldn't it be preferable to have a function to access fs_type (in
> >>>>>fs/fs.c) directly instead of a string representation?
> >>>>
> >>>>Agree, but there is no direct function in fs/fs.c.
> >>>>I'm reluctant to invent a new function just for this purpose.
> >>>
> >>>In that case we should compare to a string that is defined as constant
> >>>in fs.h.
> >
> >Are you saying that we should add new macros in fs.h?
> >All the file system names are initialized in fs/fs.c now.
> >
> >>-Takahiro Akashi
> >>
> >>>
> >>>Best regards
> >>>
> >>>Heinrich
> >>>
> >>>>
> >>>>-Takahiro Akashi
> >>>>
> >>>>
> >>>>>Otherwise we should convert the string "unsupported" of fstypes[] into a
> >>>>>constant in fs.h so that we can be sure we are using the same value.
> >>>>>
> >>>>>Best regards
> >>>>>
> >>>>>Heinrich
> >>>>>
> >>>>>>+		return 0;
> >>>>>>+
> >>>>>>+	fs_close();
> >>>>>>+
> >>>>>>+	return 1;
> >>>>>>+}
> >>>>>>+
> >>>>>>  /*
> >>>>>>   * Create a handle for a partition or disk
> >>>>>>   *
> >>>>>>@@ -270,7 +284,7 @@ static efi_status_t efi_disk_add_dev(
> >>>>>>  			       diskobj->dp);
> >>>>>>  	if (ret != EFI_SUCCESS)
> >>>>>>  		return ret;
> >>>>>>-	if (part >= 1) {
> >>>>>>+	if (part >= 1 && efi_fs_exists(desc, part)) {
> >>>>>>  		diskobj->volume = efi_simple_file_system(desc, part,
> >>>>>>  							 diskobj->dp);
> >>>>>>  		ret = efi_add_protocol(&diskobj->header,
> >>>>>>
> >>>>>
> >>>>
> >>>
> >
>