On 07.12.23 11:37, 'Michael Adler' via EFI Boot Guard wrote:
> This commit changes the default verbosity setting for the FAT parser to
> false, aligning it with the general efibootguard behavior where
> verbosity is not enabled by default. Users can still opt-in for verbose
> output by using the -v or --verbose flag.
>
This affects the output of tools only, right?
And is this a regression fix or a cleanup of an inconsistency?
> Signed-off-by: Michael Adler <michael.ad...@siemens.com>
> ---
> tools/ebgpart.c | 2 +-
> tools/fat.c | 6 +++---
> tools/fat.h | 2 +-
> tools/tests/test_fat.c | 21 +++++++++++++--------
> 4 files changed, 18 insertions(+), 13 deletions(-)
>
> diff --git a/tools/ebgpart.c b/tools/ebgpart.c
> index b6384f2..298d766 100644
> --- a/tools/ebgpart.c
> +++ b/tools/ebgpart.c
> @@ -124,7 +124,7 @@ static int check_GPT_FAT_entry(int fd, const struct
> EFIpartitionentry *e)
> strerror(errno));
> return -1;
> }
> - return determine_FAT_bits(&header);
> + return determine_FAT_bits(&header, verbosity);
> }
>
> static inline EbgFileSystemType fat_size_to_fs_type(int fat_size)
> diff --git a/tools/fat.c b/tools/fat.c
> index 0ff0f64..cc9d9ff 100644
> --- a/tools/fat.c
> +++ b/tools/fat.c
> @@ -22,8 +22,6 @@
> #include "linux_util.h"
> #include "ebgpart.h"
>
> -static bool verbosity = true;
> -
> #define fat_msg(sb, lvl, ...) do { VERBOSE(stderr, __VA_ARGS__);
> VERBOSE(stderr, "\n"); } while(0);
> #define KERN_ERR "ERROR"
>
> @@ -69,6 +67,8 @@ static int fat_read_bpb(void __attribute__((unused)) *sb,
> const struct fat_boot_sector *b, int silent,
> struct fat_bios_param_block *bpb)
> {
> + bool verbosity = !silent; /* for our VERBOSE macro */
> +
> int error = -EINVAL;
>
> /* Read in BPB ... */
> @@ -147,7 +147,7 @@ out:
> /* end of Linux kernel code */
>
> /*****************************************************************************/
>
> -int determine_FAT_bits(const struct fat_boot_sector *sector)
> +int determine_FAT_bits(const struct fat_boot_sector *sector, bool verbosity)
> {
> struct fat_bios_param_block bpb;
> if (fat_read_bpb(NULL, sector, !verbosity, &bpb)) {
> diff --git a/tools/fat.h b/tools/fat.h
> index 8c991ab..d0cf3c7 100644
> --- a/tools/fat.h
> +++ b/tools/fat.h
> @@ -22,4 +22,4 @@
> * to ensure it is a valid FAT boot sector. If the provided boot sector is
> not valid or an error
> * occurs during the determination process, the function returns a value
> less than or equal to 0.
> */
> -int determine_FAT_bits(const struct fat_boot_sector *sector);
> +int determine_FAT_bits(const struct fat_boot_sector *sector, bool verbosity);
> diff --git a/tools/tests/test_fat.c b/tools/tests/test_fat.c
> index 2c5089f..42e0133 100644
> --- a/tools/tests/test_fat.c
> +++ b/tools/tests/test_fat.c
> @@ -34,7 +34,7 @@ START_TEST(test_determine_FAT_bits_empty)
> {
> struct fat_boot_sector sector;
> memset(§or, 0, sizeof(sector));
> - int ret = determine_FAT_bits(§or);
> + int ret = determine_FAT_bits(§or, true);
> ck_assert_int_eq(ret, 0);
> }
> END_TEST
> @@ -48,7 +48,7 @@ START_TEST(test_determine_FAT_bits_sec_per_clus_zero)
> .media = 0xf8,
> };
> u16_to_le(512, sector.sector_size);
> - int ret = determine_FAT_bits(§or);
> + int ret = determine_FAT_bits(§or, true);
> ck_assert_int_eq(ret, 0);
> }
> END_TEST
> @@ -62,7 +62,7 @@ START_TEST(test_determine_FAT_bits_fat_sector_size_zero)
> .fats = 16,
> .media = 0xf8,
> };
> - int ret = determine_FAT_bits(§or);
> + int ret = determine_FAT_bits(§or, true);
> ck_assert_int_eq(ret, 0);
> }
> END_TEST
> @@ -124,7 +124,8 @@ START_TEST(test_determine_FAT_bits_12)
> 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
> 0x55, 0xaa,
> };
> - int ret = determine_FAT_bits((const struct fat_boot_sector*) §or);
> + int ret = determine_FAT_bits((const struct fat_boot_sector *)§or,
> + true);
> ck_assert_int_eq(ret, 12);
> }
> END_TEST
> @@ -186,7 +187,8 @@ START_TEST(test_determine_FAT_bits_16)
> 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
> 0x55, 0xaa,
> };
> - int ret = determine_FAT_bits((const struct fat_boot_sector*) §or);
> + int ret = determine_FAT_bits((const struct fat_boot_sector *)§or,
> + true);
> ck_assert_int_eq(ret, 16);
> }
> END_TEST
> @@ -248,7 +250,8 @@ START_TEST(test_determine_FAT_bits_32)
> 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
> 0x55, 0xaa,
> };
> - int ret = determine_FAT_bits((const struct fat_boot_sector*) §or);
> + int ret = determine_FAT_bits((const struct fat_boot_sector *)§or,
> + true);
> ck_assert_int_eq(ret, 32);
> }
> END_TEST
> @@ -312,7 +315,8 @@ START_TEST(test_determine_FAT_bits_fat16_swupdate)
> 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
> 0x55, 0xaa,
> };
> - int ret = determine_FAT_bits((const struct fat_boot_sector*) §or);
> + int ret = determine_FAT_bits((const struct fat_boot_sector *)§or,
> + true);
> ck_assert_int_eq(ret, 16);
> }
> END_TEST
> @@ -373,7 +377,8 @@ START_TEST(test_determine_FAT_bits_squashfs)
> 0xd7, 0x40, 0xc1, 0xcc, 0x43, 0x7b, 0xbf, 0x8d, 0x76, 0x39,
> 0x3c, 0xd5,
> };
> - int ret = determine_FAT_bits((const struct fat_boot_sector *)§or);
> + int ret = determine_FAT_bits((const struct fat_boot_sector *)§or,
> + true);
> ck_assert_int_eq(ret, 0);
> }
> END_TEST
Looks good otherwise.
Jan
--
Siemens AG, Technology
Linux Expert Center
--
You received this message because you are subscribed to the Google Groups "EFI
Boot Guard" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to efibootguard-dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/efibootguard-dev/abcc4201-a8d8-4b0d-b3fe-63ee849ad15f%40siemens.com.
