Am 18. September 2026 22:35:33 MESZ schrieb Han Gao <[email protected]>:
>Storage capacities above 4 GiB are truncated when stored in size_t
>on 32-bit systems. Use a 64-bit accumulator and promote the block
>count before multiplication so 32-bit LBA configurations are also
>handled correctly.
>
>Fixes: 86d462c05d57 ("cmd: add a fetch utility")
>
>Signed-off-by: Han Gao <[email protected]>
>---
> cmd/ufetch.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
>diff --git a/cmd/ufetch.c b/cmd/ufetch.c
>index d4b54f77b7b2..29ccbf654a4e 100644
>--- a/cmd/ufetch.c
>+++ b/cmd/ufetch.c
>@@ -95,7 +95,7 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int
>argc,
> const char *model, *compatible;
> char *ipaddr;
> int n_cmds, n_cpus = 0, compatlen;
>- size_t size = 0;
>+ u64 size = 0;
> ofnode np;
> bool skip_ascii = false;
>
>@@ -221,7 +221,7 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int
>argc,
> ret = uclass_find_device_by_seq(UCLASS_BLK, line -
> STORAGE, &dev);
> if (!ret && dev) {
> desc = dev_get_uclass_plat(dev);
>- size = desc->lba * desc->blksz;
>+ size = (u64)desc->lba * desc->blksz;
Shouldn't desc->lba itself be 64 bit to allow for drives larger than 2 TiB.
Why do we allow CONFIG_SYS_64BIT_LBA=n?
Best regards
Heinrich
> printf("%4s %d: " RESET,
> blk_get_uclass_name(desc->uclass_id),
> desc->lun);
> if (size)
>
>---
>base-commit: 211de43d0f954a00a490220c1aac9db298287c40
>change-id: 20260919-fixfetch-d7aa5b94d9b7
>
>Best regards,
>--
>Han Gao <[email protected]>
>
