Re: [PULL 00/13] Net patches

[Peter Maydell]

On Fri, 27 Mar 2020 at 11:14, Jason Wang <address@hidden> wrote:
>
> The following changes since commit cfe68ae025f704f336d7dd3d1903ce37b445831d:
>
>   Merge remote-tracking branch 
> 'remotes/vivier2/tags/linux-user-for-5.0-pull-request' into staging 
> (2020-03-26 20:55:54 +0000)
>
> are available in the git repository at:
>
>   https://github.com/jasowang/qemu.git tags/net-pull-request
>
> for you to fetch changes up to f3b364f4f77fcb24cec468f518bf5e093dc27cb7:
>
>   hw/net/allwinner-sun8i-emac.c: Fix REG_ADDR_HIGH/LOW reads (2020-03-27 
> 18:59:47 +0800)
>
> ----------------------------------------------------------------
>
> ----------------------------------------------------------------

Hi; this fails to compile (all platforms):

/home/petmay01/qemu-for-merges/hw/net/allwinner-sun8i-emac.c:773:20:
error: initialization from incompatible pointer type
[-Werror=incompatible-pointer-types]
     .can_receive = allwinner_sun8i_emac_can_receive,
                    ^
/home/petmay01/qemu-for-merges/hw/net/allwinner-sun8i-emac.c:773:20:
note: (near initialization for
'net_allwinner_sun8i_emac_info.can_receive')


There's also this one, though not every compiler picked it up:

/home/peter.maydell/qemu/hw/net/i82596.c: In function 'i82596_receive':
/home/peter.maydell/qemu/hw/net/i82596.c:657:30: error: comparison of
unsigned expression >= 0 is always true [-Werror=type-limits]
                 assert(bufsz >= 0);
                              ^
/home/peter.maydell/qemu/hw/net/i82596.c:657:30: error: comparison of
unsigned expression >= 0 is always true [-Werror=type-limits]
                 assert(bufsz >= 0);
                              ^


For the first error, I think this needs squashing into
"hw/net: Make NetCanReceive() return a boolean":

diff --git a/hw/net/allwinner-sun8i-emac.c b/hw/net/allwinner-sun8i-emac.c
index fc67a1be70..28637ff4c1 100644
--- a/hw/net/allwinner-sun8i-emac.c
+++ b/hw/net/allwinner-sun8i-emac.c
@@ -395,7 +395,7 @@ static void
allwinner_sun8i_emac_flush_desc(FrameDescriptor *desc,
     cpu_physical_memory_write(phys_addr, desc, sizeof(*desc));
 }

-static int allwinner_sun8i_emac_can_receive(NetClientState *nc)
+static bool allwinner_sun8i_emac_can_receive(NetClientState *nc)
 {
     AwSun8iEmacState *s = qemu_get_nic_opaque(nc);
     FrameDescriptor desc;


Squashing this into my
"hw/net/i82596.c: Avoid reading off end of buffer in i82596_receive()"
commit fixes the second error.

diff --git a/hw/net/i82596.c b/hw/net/i82596.c
index a9bdbac339..055c3a1470 100644
--- a/hw/net/i82596.c
+++ b/hw/net/i82596.c
@@ -653,8 +653,8 @@ ssize_t i82596_receive(NetClientState *nc, const
uint8_t *buf, size_t sz)

             if (bufcount > 0) {
                 /* Still some of the actual data buffer to transfer */
+                assert(bufsz >= bufcount);
                 bufsz -= bufcount;
-                assert(bufsz >= 0);
                 address_space_write(&address_space_memory, rba,
                                     MEMTXATTRS_UNSPECIFIED, buf, bufcount);
                 rba += bufcount;

thanks
-- PMM