On Mon, May 22, 2023 at 02:04:41PM -0500, Eric Blake wrote:
> We have several limitations and bugs worth fixing; they are
> inter-related enough that it is not worth splitting this patch into
> smaller pieces:
>
> +++ b/util/cutils.c
> @@ -194,15 +194,18 @@ static int64_t suffix_mul(char suffix, int64_t unit)
> * - 12345 - decimal, scale determined by @default_suffix and @unit
> * - 12345{bBkKmMgGtTpPeE} - decimal, scale determined by suffix and @unit
> * - 12345.678{kKmMgGtTpPeE} - decimal, scale determined by suffix, and
> - * fractional portion is truncated to byte
> + * fractional portion is truncated to byte, either side of . may be empty
> * - 0x7fEE - hexadecimal, unit determined by @default_suffix
> *
> * The following are intentionally not supported
> - * - hex with scaling suffix, such as 0x20M
> - * - octal, such as 08
> - * - fractional hex, such as 0x1.8
> - * - negative values, including -0
> - * - floating point exponents, such as 1e3
> + * - hex with scaling suffix, such as 0x20M (0x1b is 27, not 1)
> + * - octal, such as 08 (parsed as decimal instead)
> + * - binary, such as 0b1000 (parsed as 0b with trailing garbage "1000")
> + * - fractional hex, such as 0x1.8 (parsed as 0 with trailing garbage "x1.8")
> + * - negative values, including -0 (fail with -ERANGE)
> + * - floating point exponents, such as 1e3 (parsed as 1e with trailing
> + * garbage "3") or 0x1p3 (parsed as 1 with trailing garbage "p3")
This latter clause is wrong - we reject 0x1p3 earlier under the hex
with scaling suffix rule. I've touched up the comment as part of
preparing the pull request.
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3266
Virtualization: qemu.org | libvirt.org
