On Fri, Aug 22, 2025 at 02:26:42PM +0200, Paolo Bonzini wrote:
> Date: Fri, 22 Aug 2025 14:26:42 +0200
> From: Paolo Bonzini <[email protected]>
> Subject: [PATCH 01/14] treewide: write "unsigned long int" instead of "long
> unsigned int"
> X-Mailer: git-send-email 2.50.1
>
> Putting "unsigned" in anything but the first position is weird.
I think one reason may be gcc uses something like ‘long unsigned int *‘
by default?
../hw/misc/imx7_src.c: In function ‘imx7_src_write’:
../hw/misc/imx7_src.c:218:42: error: passing argument 2 of ‘clear_bit’ from
incompatible pointer type [-Werror=incompatible-pointer-types]
218 | clear_bit(R_CORE1_RST_SHIFT, &change_mask);
| ^~~~~~~~~~~~
| |
| uint32_t * {aka unsigned int *}
In file included from /media/liuzhao/data/qemu-cook/include/qemu/bitmap.h:16,
from /media/liuzhao/data/qemu-cook/include/hw/qdev-core.h:6,
from /media/liuzhao/data/qemu-cook/include/hw/sysbus.h:6,
from
/media/liuzhao/data/qemu-cook/include/hw/misc/imx7_src.h:13,
from ../hw/misc/imx7_src.c:12:
/qemu/include/qemu/bitops.h:93:54: note: expected ‘long unsigned int *’ but
argument is of type ‘uint32_t *’ {aka ‘unsigned int *’}
93 | static inline void clear_bit(long nr, unsigned long *addr)
| ~~~~~~~~~~~~~~~^~~~
cc1: all warnings being treated as errors
> As such,
> tracetool's Rust type conversion will not support it. Remove it from
> the whole of QEMU's source code, not just trace-events.
But I also agree it's a good idea to clean this up.
> Signed-off-by: Paolo Bonzini <[email protected]>
> ---
> crypto/pbkdf-gcrypt.c | 2 +-
> crypto/pbkdf-gnutls.c | 2 +-
> crypto/pbkdf-nettle.c | 2 +-
> hw/display/exynos4210_fimd.c | 2 +-
> hw/misc/imx7_src.c | 4 ++--
> hw/net/can/can_sja1000.c | 4 ++--
> hw/xen/trace-events | 4 ++--
> 7 files changed, 10 insertions(+), 10 deletions(-)
...
> diff --git a/hw/misc/imx7_src.c b/hw/misc/imx7_src.c
> index df0b0a69057..817c95bf65b 100644
> --- a/hw/misc/imx7_src.c
> +++ b/hw/misc/imx7_src.c
> @@ -169,7 +169,7 @@ static void imx7_src_write(void *opaque, hwaddr offset,
> uint64_t value,
> {
> IMX7SRCState *s = (IMX7SRCState *)opaque;
> uint32_t index = offset >> 2;
> - long unsigned int change_mask;
> + uint32_t change_mask;
We needs "unsigned long", otherwise, there'll be the error as I listed
above.
> uint32_t current_value = value;
>
> if (index >= SRC_MAX) {
...
> diff --git a/hw/net/can/can_sja1000.c b/hw/net/can/can_sja1000.c
> index 5b6ba9df6c4..545c520c3b4 100644
> --- a/hw/net/can/can_sja1000.c
> +++ b/hw/net/can/can_sja1000.c
> @@ -750,8 +750,8 @@ uint64_t can_sja_mem_read(CanSJA1000State *s, hwaddr
> addr, unsigned size)
> break;
> }
> }
> - DPRINTF("read addr 0x%02x, %d bytes, content 0x%02lx\n",
> - (int)addr, size, (long unsigned int)temp);
tmep is "uint64_t", so there's no need to convert its type?
We can just drop `(long unsigned int)` directly.
> + DPRINTF("read addr 0x%02x, %d bytes, content 0x%02x\n",
> + (int)addr, size, (unsigned)temp);
> return temp;
> }
Others look fine to me. With the nits fixed,
Reviewed-by: Zhao Liu <[email protected]>
