18.08.2021 08:25, Christoph Hellwig пишет:
> On Wed, Aug 18, 2021 at 03:55:47AM +0300, Dmitry Osipenko wrote:
>> Support looking up GPT at a non-standard location specified by a block
>> device driver.
>>
>> Signed-off-by: Dmitry Osipenko <digetx@xxxxxxxxx>
>> ---
>> block/partitions/efi.c | 13 +++++++++++++
>> 1 file changed, 13 insertions(+)
>>
>> diff --git a/block/partitions/efi.c b/block/partitions/efi.c
>> index aaa3dc487cb5..b9509f445b3c 100644
>> --- a/block/partitions/efi.c
>> +++ b/block/partitions/efi.c
>> @@ -585,6 +585,8 @@ static int find_valid_gpt(struct parsed_partitions *state, gpt_header **gpt,
>> gpt_header *pgpt = NULL, *agpt = NULL;
>> gpt_entry *pptes = NULL, *aptes = NULL;
>> legacy_mbr *legacymbr;
>> + struct gendisk *disk = state->disk;
>> + const struct block_device_operations *fops = disk->fops;
>> sector_t total_sectors = get_capacity(state->disk);
>> u64 lastlba;
>>
>> @@ -619,6 +621,17 @@ static int find_valid_gpt(struct parsed_partitions *state, gpt_header **gpt,
>> if (!good_agpt && force_gpt)
>> good_agpt = is_gpt_valid(state, lastlba, &agpt, &aptes);
>>
>> + if (!good_agpt && force_gpt && fops->alternative_gpt_sector) {
>> + struct block_device *bdev = disk->part0;
>> + sector_t agpt_sector;
>> + int err;
>> +
>> + err = fops->alternative_gpt_sector(bdev, &agpt_sector);
>
> Please call the method with the disk as the argument. I've been moving
> the block layer to generally pass the gendisk whenever we're dealing
> with the whole device, as that makes the intent very clear.
I'll change it in v6.
> Also do we really need the force_gpt check? That is we always require
> the user to pass a command line argument to use this?
User indeed must pass the 'gpt' argument to kernel cmdline. That's what
all those Android devices do. Should be okay to keep that requirement.
