[PATCH v4 0/7] linux-user: Add support for various missing netlink sockopt entries

[PATCH v4 0/7] linux-user: Add support for various missing netlink sockopt entries

[deller]

From: Helge Deller <deller@gmx.de>

This patchset adds various missing sockopt calls, so that qemu linux-user
is able to successfully build the debian gupnp package in a chroot.

Tested with a 32-bit big-endian hppa linux-user chroot running on a phyiscal
x86-64 little-endian host.

This fixes debian's bug report:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1044651

Signed-off-by: Helge Deller <deller@gmx.de>

Please review patches #1 and #7.

Changes in v4:
- rewrote patch #1 to fix build on ubuntu 22.04 and add
  cleanup-patch #7
- added S-o-b from Laurent to patches #3 and #4

Changes in v3:
- Updated patches #3, #4 and #5 based on suggestions from
  Laurent Vivier

Changes in v2:
- Moved IFA_PROTO in patch 1 as suggested by Laurent Vivier
- Added Reviewed-by from Philippe Mathieu-Daudé and Laurent Vivier
  in patch 2

Helge Deller (7):
  linux-user: netlink: Add missing IFA_PROTO to
    host_to_target_data_addr_rtattr()
  linux-user: Use unique error messages for cmsg parsing
  linux-user: netlink: Add IP_PKTINFO cmsg parsing
  linux-user: netlink: Add emulation of IP_MULTICAST_IF
  linux-user: netlink: add netlink neighbour emulation
  linux-user: netlink: Add missing QEMU_IFLA entries
  linux-user: netlink: Use QEMU_IFA_XXX values instead of IFA_XXX

 linux-user/fd-trans.c     | 169 ++++++++++++++++++++++++++++++++++++--
 linux-user/syscall.c      |  34 ++++++--
 linux-user/syscall_defs.h |   6 ++
 3 files changed, 192 insertions(+), 17 deletions(-)

-- 
2.47.0

[PATCH v4 1/7] linux-user: netlink: Add missing IFA_PROTO to host_to_target_data_addr_rtattr()

[deller]

From: Helge Deller <deller@gmx.de>

Fixes this warning:
 Unknown host IFA type: 11

IFA_PROTO has been introduced in kernel v5.18, and as such using it
unconditionally breaks build on Ubuntu 22.04.  Instead convert the IFA_XXX
values to QEMU_IDA_XXX values and use those instead.
Already existing IFA_XXX usages are converted to QEMU_IFA_XXX in a
followup-patch.

Signed-off-by: Helge Deller <deller@gmx.de>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
---
 linux-user/fd-trans.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
index c04a97c73a..05a944b32e 100644
--- a/linux-user/fd-trans.c
+++ b/linux-user/fd-trans.c
@@ -31,6 +31,22 @@
 #include "fd-trans.h"
 #include "signal-common.h"
 
+enum {
+    QEMU_IFA_UNSPEC,
+    QEMU_IFA_ADDRESS,
+    QEMU_IFA_LOCAL,
+    QEMU_IFA_LABEL,
+    QEMU_IFA_BROADCAST,
+    QEMU_IFA_ANYCAST,
+    QEMU_IFA_CACHEINFO,
+    QEMU_IFA_MULTICAST,
+    QEMU_IFA_FLAGS,
+    QEMU_IFA_RT_PRIORITY,
+    QEMU_IFA_TARGET_NETNSID,
+    QEMU_IFA_PROTO,
+    QEMU__IFA__MAX,
+};
+
 enum {
     QEMU_IFLA_BR_UNSPEC,
     QEMU_IFLA_BR_FORWARD_DELAY,
@@ -1140,6 +1156,7 @@ static abi_long host_to_target_data_addr_rtattr(struct rtattr *rtattr)
     /* binary: depends on family type */
     case IFA_ADDRESS:
     case IFA_LOCAL:
+    case QEMU_IFA_PROTO:
         break;
     /* string */
     case IFA_LABEL:
-- 
2.47.0

Re: [PATCH v4 1/7] linux-user: netlink: Add missing IFA_PROTO to host_to_target_data_addr_rtattr()

[Laurent Vivier]

Le 20/01/2025 à 22:33, deller@kernel.org a écrit :
> From: Helge Deller <deller@gmx.de>
> 
> Fixes this warning:
>   Unknown host IFA type: 11
> 
> IFA_PROTO has been introduced in kernel v5.18, and as such using it
> unconditionally breaks build on Ubuntu 22.04.  Instead convert the IFA_XXX
> values to QEMU_IDA_XXX values and use those instead.
> Already existing IFA_XXX usages are converted to QEMU_IFA_XXX in a
> followup-patch.
> 
> Signed-off-by: Helge Deller <deller@gmx.de>
> Reviewed-by: Laurent Vivier <laurent@vivier.eu>
> ---
>   linux-user/fd-trans.c | 17 +++++++++++++++++
>   1 file changed, 17 insertions(+)
> 
> diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
> index c04a97c73a..05a944b32e 100644
> --- a/linux-user/fd-trans.c
> +++ b/linux-user/fd-trans.c
> @@ -31,6 +31,22 @@
>   #include "fd-trans.h"
>   #include "signal-common.h"
>   
> +enum {
> +    QEMU_IFA_UNSPEC,
> +    QEMU_IFA_ADDRESS,
> +    QEMU_IFA_LOCAL,
> +    QEMU_IFA_LABEL,
> +    QEMU_IFA_BROADCAST,
> +    QEMU_IFA_ANYCAST,
> +    QEMU_IFA_CACHEINFO,
> +    QEMU_IFA_MULTICAST,
> +    QEMU_IFA_FLAGS,
> +    QEMU_IFA_RT_PRIORITY,
> +    QEMU_IFA_TARGET_NETNSID,
> +    QEMU_IFA_PROTO,
> +    QEMU__IFA__MAX,
> +};
> +
>   enum {
>       QEMU_IFLA_BR_UNSPEC,
>       QEMU_IFLA_BR_FORWARD_DELAY,
> @@ -1140,6 +1156,7 @@ static abi_long host_to_target_data_addr_rtattr(struct rtattr *rtattr)
>       /* binary: depends on family type */
>       case IFA_ADDRESS:
>       case IFA_LOCAL:
> +    case QEMU_IFA_PROTO:
>           break;
>       /* string */
>       case IFA_LABEL:

It would be cleaner to replace all the IFA_XXX by their QEMU_IFA_XXX.

Thanks,
Laurent

Re: [PATCH v4 1/7] linux-user: netlink: Add missing IFA_PROTO to host_to_target_data_addr_rtattr()

[Helge Deller]

* Laurent Vivier <laurent@vivier.eu>:
> [...]
> It would be cleaner to replace all the IFA_XXX by their QEMU_IFA_XXX.

Thanks for review, Laurent!

Below I've merged patch #1 and #7, as suggested by you.
Is this OK?
Can you review?

Thanks!
Helge
________________________________________

From: Helge Deller <deller@gmx.de>
Date: Mon, 20 Jan 2025 22:20:41 +0100
Subject: [PATCH] linux-user: netlink: Add missing IFA_PROTO to
 host_to_target_data_addr_rtattr()

Fix this warning:
 Unknown host IFA type: 11

While adding IFA_PROTO, convert all IFA_XXX values over to QEMU_IFA_XXX values
to avoid a build failure on Ubuntu 22.04 (kernel v5.18 which does not know
IFA_PROTO yet).

Signed-off-by: Helge Deller <deller@gmx.de>

diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
index c04a97c73a..2e714c8e56 100644
--- a/linux-user/fd-trans.c
+++ b/linux-user/fd-trans.c
@@ -31,6 +31,22 @@
 #include "fd-trans.h"
 #include "signal-common.h"
 
+enum {
+    QEMU_IFA_UNSPEC,
+    QEMU_IFA_ADDRESS,
+    QEMU_IFA_LOCAL,
+    QEMU_IFA_LABEL,
+    QEMU_IFA_BROADCAST,
+    QEMU_IFA_ANYCAST,
+    QEMU_IFA_CACHEINFO,
+    QEMU_IFA_MULTICAST,
+    QEMU_IFA_FLAGS,
+    QEMU_IFA_RT_PRIORITY,
+    QEMU_IFA_TARGET_NETNSID,
+    QEMU_IFA_PROTO,
+    QEMU__IFA__MAX,
+};
+
 enum {
     QEMU_IFLA_BR_UNSPEC,
     QEMU_IFLA_BR_FORWARD_DELAY,
@@ -1138,20 +1154,21 @@ static abi_long host_to_target_data_addr_rtattr(struct rtattr *rtattr)
 
     switch (rtattr->rta_type) {
     /* binary: depends on family type */
-    case IFA_ADDRESS:
-    case IFA_LOCAL:
+    case QEMU_IFA_ADDRESS:
+    case QEMU_IFA_LOCAL:
+    case QEMU_IFA_PROTO:
         break;
     /* string */
-    case IFA_LABEL:
+    case QEMU_IFA_LABEL:
         break;
     /* u32 */
-    case IFA_FLAGS:
-    case IFA_BROADCAST:
+    case QEMU_IFA_FLAGS:
+    case QEMU_IFA_BROADCAST:
         u32 = RTA_DATA(rtattr);
         *u32 = tswap32(*u32);
         break;
     /* struct ifa_cacheinfo */
-    case IFA_CACHEINFO:
+    case QEMU_IFA_CACHEINFO:
         ci = RTA_DATA(rtattr);
         ci->ifa_prefered = tswap32(ci->ifa_prefered);
         ci->ifa_valid = tswap32(ci->ifa_valid);
@@ -1398,8 +1415,8 @@ static abi_long target_to_host_data_addr_rtattr(struct rtattr *rtattr)
 {
     switch (rtattr->rta_type) {
     /* binary: depends on family type */
-    case IFA_LOCAL:
-    case IFA_ADDRESS:
+    case QEMU_IFA_LOCAL:
+    case QEMU_IFA_ADDRESS:
         break;
     default:
         qemu_log_mask(LOG_UNIMP, "Unknown target IFA type: %d\n",

Re: [PATCH v4 1/7] linux-user: netlink: Add missing IFA_PROTO to host_to_target_data_addr_rtattr()

[Laurent Vivier]

Le 22/01/2025 à 22:20, Helge Deller a écrit :
> * Laurent Vivier <laurent@vivier.eu>:
>> [...]
>> It would be cleaner to replace all the IFA_XXX by their QEMU_IFA_XXX.
> 
> Thanks for review, Laurent!
> 
> Below I've merged patch #1 and #7, as suggested by you.
> Is this OK?

Yes, thanks
> Can you review?

Reviewed-by: Laurent Vivier <laurent@vivier.eu>

> 
> Thanks!
> Helge
> ________________________________________
> 
> From: Helge Deller <deller@gmx.de>
> Date: Mon, 20 Jan 2025 22:20:41 +0100
> Subject: [PATCH] linux-user: netlink: Add missing IFA_PROTO to
>   host_to_target_data_addr_rtattr()
> 
> Fix this warning:
>   Unknown host IFA type: 11
> 
> While adding IFA_PROTO, convert all IFA_XXX values over to QEMU_IFA_XXX values
> to avoid a build failure on Ubuntu 22.04 (kernel v5.18 which does not know
> IFA_PROTO yet).
> 
> Signed-off-by: Helge Deller <deller@gmx.de>
> 
> diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
> index c04a97c73a..2e714c8e56 100644
> --- a/linux-user/fd-trans.c
> +++ b/linux-user/fd-trans.c
> @@ -31,6 +31,22 @@
>   #include "fd-trans.h"
>   #include "signal-common.h"
>   
> +enum {
> +    QEMU_IFA_UNSPEC,
> +    QEMU_IFA_ADDRESS,
> +    QEMU_IFA_LOCAL,
> +    QEMU_IFA_LABEL,
> +    QEMU_IFA_BROADCAST,
> +    QEMU_IFA_ANYCAST,
> +    QEMU_IFA_CACHEINFO,
> +    QEMU_IFA_MULTICAST,
> +    QEMU_IFA_FLAGS,
> +    QEMU_IFA_RT_PRIORITY,
> +    QEMU_IFA_TARGET_NETNSID,
> +    QEMU_IFA_PROTO,
> +    QEMU__IFA__MAX,
> +};
> +
>   enum {
>       QEMU_IFLA_BR_UNSPEC,
>       QEMU_IFLA_BR_FORWARD_DELAY,
> @@ -1138,20 +1154,21 @@ static abi_long host_to_target_data_addr_rtattr(struct rtattr *rtattr)
>   
>       switch (rtattr->rta_type) {
>       /* binary: depends on family type */
> -    case IFA_ADDRESS:
> -    case IFA_LOCAL:
> +    case QEMU_IFA_ADDRESS:
> +    case QEMU_IFA_LOCAL:
> +    case QEMU_IFA_PROTO:
>           break;
>       /* string */
> -    case IFA_LABEL:
> +    case QEMU_IFA_LABEL:
>           break;
>       /* u32 */
> -    case IFA_FLAGS:
> -    case IFA_BROADCAST:
> +    case QEMU_IFA_FLAGS:
> +    case QEMU_IFA_BROADCAST:
>           u32 = RTA_DATA(rtattr);
>           *u32 = tswap32(*u32);
>           break;
>       /* struct ifa_cacheinfo */
> -    case IFA_CACHEINFO:
> +    case QEMU_IFA_CACHEINFO:
>           ci = RTA_DATA(rtattr);
>           ci->ifa_prefered = tswap32(ci->ifa_prefered);
>           ci->ifa_valid = tswap32(ci->ifa_valid);
> @@ -1398,8 +1415,8 @@ static abi_long target_to_host_data_addr_rtattr(struct rtattr *rtattr)
>   {
>       switch (rtattr->rta_type) {
>       /* binary: depends on family type */
> -    case IFA_LOCAL:
> -    case IFA_ADDRESS:
> +    case QEMU_IFA_LOCAL:
> +    case QEMU_IFA_ADDRESS:
>           break;
>       default:
>           qemu_log_mask(LOG_UNIMP, "Unknown target IFA type: %d\n",

[PATCH v4 2/7] linux-user: Use unique error messages for cmsg parsing

[deller]

From: Helge Deller <deller@gmx.de>

Avoid using the same error message for two different code paths
as it complicates determining the one which actually triggered.

Signed-off-by: Helge Deller <deller@gmx.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
---
 linux-user/syscall.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index 78c7c0b34e..a157abc40c 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -1827,7 +1827,7 @@ static inline abi_long target_to_host_cmsg(struct msghdr *msgh,
                 *dst = tswap32(*dst);
             }
         } else {
-            qemu_log_mask(LOG_UNIMP, "Unsupported ancillary data: %d/%d\n",
+            qemu_log_mask(LOG_UNIMP, "Unsupported target ancillary data: %d/%d\n",
                           cmsg->cmsg_level, cmsg->cmsg_type);
             memcpy(data, target_data, len);
         }
@@ -2049,7 +2049,7 @@ static inline abi_long host_to_target_cmsg(struct target_msghdr *target_msgh,
 
         default:
         unimplemented:
-            qemu_log_mask(LOG_UNIMP, "Unsupported ancillary data: %d/%d\n",
+            qemu_log_mask(LOG_UNIMP, "Unsupported host ancillary data: %d/%d\n",
                           cmsg->cmsg_level, cmsg->cmsg_type);
             memcpy(target_data, data, MIN(len, tgt_len));
             if (tgt_len > len) {
-- 
2.47.0

[PATCH v4 3/7] linux-user: netlink: Add IP_PKTINFO cmsg parsing

[deller]

From: Helge Deller <deller@gmx.de>

Fixes those warnings:
 Unsupported host ancillary data: 0/8

Signed-off-by: Helge Deller <deller@gmx.de>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
---
 linux-user/syscall.c      | 10 ++++++++++
 linux-user/syscall_defs.h |  6 ++++++
 2 files changed, 16 insertions(+)

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index a157abc40c..df8609b4d8 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -1998,6 +1998,16 @@ static inline abi_long host_to_target_cmsg(struct target_msghdr *target_msgh,
                     (void *) &errh->offender, sizeof(errh->offender));
                 break;
             }
+            case IP_PKTINFO:
+            {
+                struct in_pktinfo *pkti = data;
+                struct target_in_pktinfo *target_pi = target_data;
+
+                __put_user(pkti->ipi_ifindex, &target_pi->ipi_ifindex);
+                target_pi->ipi_spec_dst.s_addr = pkti->ipi_spec_dst.s_addr;
+                target_pi->ipi_addr.s_addr = pkti->ipi_addr.s_addr;
+                break;
+            }
             default:
                 goto unimplemented;
             }
diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
index faad9147c9..86d773add7 100644
--- a/linux-user/syscall_defs.h
+++ b/linux-user/syscall_defs.h
@@ -2622,6 +2622,12 @@ struct target_ucred {
     abi_uint gid;
 };
 
+struct target_in_pktinfo {
+    abi_int               ipi_ifindex;
+    struct target_in_addr ipi_spec_dst;
+    struct target_in_addr ipi_addr;
+};
+
 typedef abi_int target_timer_t;
 
 #define TARGET_SIGEV_MAX_SIZE 64
-- 
2.47.0

[PATCH v4 4/7] linux-user: netlink: Add emulation of IP_MULTICAST_IF

[deller]

From: Helge Deller <deller@gmx.de>

Add IP_MULTICAST_IF and share the code with IP_ADD_MEMBERSHIP / IP_DROP_MEMBERSHIP.
Sharing the code makes sense, because the manpage of ip(7) says:

IP_MULTICAST_IF (since Linux 1.2)
      Set the local device for a multicast socket.  The argument
      for setsockopt(2) is an ip_mreqn or (since Linux 3.5)
      ip_mreq structure similar to IP_ADD_MEMBERSHIP, or an
      in_addr structure.  (The kernel determines which structure
      is being passed based on the size passed in optlen.)  For
      getsockopt(2), the argument is an in_addr structure.

Signed-off-by: Helge Deller <deller@gmx.de>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
---
 linux-user/syscall.c | 20 ++++++++++++++------
 1 file changed, 14 insertions(+), 6 deletions(-)

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index df8609b4d8..6ee02383da 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -2130,16 +2130,23 @@ static abi_long do_setsockopt(int sockfd, int level, int optname,
             }
             ret = get_errno(setsockopt(sockfd, level, optname, &val, sizeof(val)));
             break;
+        case IP_MULTICAST_IF:
         case IP_ADD_MEMBERSHIP:
         case IP_DROP_MEMBERSHIP:
         {
             struct ip_mreqn ip_mreq;
             struct target_ip_mreqn *target_smreqn;
+            int min_size;
 
             QEMU_BUILD_BUG_ON(sizeof(struct ip_mreq) !=
                               sizeof(struct target_ip_mreq));
 
-            if (optlen < sizeof (struct target_ip_mreq) ||
+            if (optname == IP_MULTICAST_IF) {
+                min_size = sizeof(struct in_addr);
+            } else {
+                min_size = sizeof(struct target_ip_mreq);
+            }
+            if (optlen < min_size ||
                 optlen > sizeof (struct target_ip_mreqn)) {
                 return -TARGET_EINVAL;
             }
@@ -2149,13 +2156,14 @@ static abi_long do_setsockopt(int sockfd, int level, int optname,
                 return -TARGET_EFAULT;
             }
             ip_mreq.imr_multiaddr.s_addr = target_smreqn->imr_multiaddr.s_addr;
-            ip_mreq.imr_address.s_addr = target_smreqn->imr_address.s_addr;
-            if (optlen == sizeof(struct target_ip_mreqn)) {
-                ip_mreq.imr_ifindex = tswapal(target_smreqn->imr_ifindex);
-                optlen = sizeof(struct ip_mreqn);
+            if (optlen >= sizeof(struct target_ip_mreq)) {
+                ip_mreq.imr_address.s_addr = target_smreqn->imr_address.s_addr;
+                if (optlen >= sizeof(struct target_ip_mreqn)) {
+                    __put_user(target_smreqn->imr_ifindex, &ip_mreq.imr_ifindex);
+                    optlen = sizeof(struct ip_mreqn);
+                }
             }
             unlock_user(target_smreqn, optval_addr, 0);
-
             ret = get_errno(setsockopt(sockfd, level, optname, &ip_mreq, optlen));
             break;
         }
-- 
2.47.0

[PATCH v4 5/7] linux-user: netlink: add netlink neighbour emulation

[deller]

From: Helge Deller <deller@gmx.de>

Fixes various warnings in the testsuite while building gupnp:
 gssdp-net-DEBUG: Failed to send netlink message: Operation not supported
 gupnp-context-DEBUG: Mismatch between host header and host IP (example.com, expected: 127.0.0.1)
 gupnp-context-DEBUG: Mismatch between host header and host port (80, expected 4711)
 gupnp-context-DEBUG: Mismatch between host header and host IP (192.168.1.2, expected: 127.0.0.1)
 gupnp-context-DEBUG: Mismatch between host header and host IP (fe80::01, expected: 127.0.0.1)
 gupnp-context-DEBUG: Mismatch between host header and host port (80, expected 4711)
 gupnp-context-DEBUG: Failed to parse HOST header from request: Invalid IPv6 address ?[fe80::01%1]? in URI
 gupnp-context-DEBUG: Failed to parse HOST header from request: Invalid IPv6 address ?[fe80::01%eth0]? in URI
 gupnp-context-DEBUG: Failed to parse HOST header from request: Could not parse port ?:1? in URI
 gupnp-context-DEBUG: Mismatch between host header and host IP (example.com, expected: ::1)
 gupnp-context-DEBUG: Mismatch between host header and host port (80, expected 4711)
 gupnp-context-DEBUG: Mismatch between host header and host IP (example.com, expected: ::1)
 gupnp-context-DEBUG: Mismatch between host header and host port (80, expected 4711)
 gupnp-context-DEBUG: Mismatch between host header and host IP (example.com, expected: ::1)

Signed-off-by: Helge Deller <deller@gmx.de>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
---
 linux-user/fd-trans.c | 100 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 100 insertions(+)

diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
index 05a944b32e..59b622aba9 100644
--- a/linux-user/fd-trans.c
+++ b/linux-user/fd-trans.c
@@ -25,12 +25,16 @@
 #ifdef CONFIG_RTNETLINK
 #include <linux/rtnetlink.h>
 #include <linux/if_bridge.h>
+#include <linux/neighbour.h>
 #endif
 #include "qemu.h"
 #include "user-internals.h"
 #include "fd-trans.h"
 #include "signal-common.h"
 
+#define NDM_RTA(r)  ((struct rtattr*)(((char*)(r)) + \
+                    NLMSG_ALIGN(sizeof(struct ndmsg))))
+
 enum {
     QEMU_IFA_UNSPEC,
     QEMU_IFA_ADDRESS,
@@ -1226,6 +1230,35 @@ static abi_long host_to_target_data_route_rtattr(struct rtattr *rtattr)
     return 0;
 }
 
+static abi_long host_to_target_data_neigh_rtattr(struct rtattr *rtattr)
+{
+    struct nda_cacheinfo *ndac;
+    uint32_t *u32;
+
+    switch (rtattr->rta_type) {
+    case NDA_UNSPEC:
+    case NDA_DST:
+    case NDA_LLADDR:
+        break;
+    case NDA_PROBES:
+        u32 = RTA_DATA(rtattr);
+        *u32 = tswap32(*u32);
+        break;
+    case NDA_CACHEINFO:
+        ndac = RTA_DATA(rtattr);
+        ndac->ndm_confirmed = tswap32(ndac->ndm_confirmed);
+        ndac->ndm_used      = tswap32(ndac->ndm_used);
+        ndac->ndm_updated   = tswap32(ndac->ndm_updated);
+        ndac->ndm_refcnt    = tswap32(ndac->ndm_refcnt);
+        break;
+    default:
+        qemu_log_mask(LOG_UNIMP, "Unknown host to target NEIGH type: %d\n",
+                      rtattr->rta_type);
+        break;
+    }
+    return 0;
+}
+
 static abi_long host_to_target_link_rtattr(struct rtattr *rtattr,
                                          uint32_t rtattr_len)
 {
@@ -1247,12 +1280,20 @@ static abi_long host_to_target_route_rtattr(struct rtattr *rtattr,
                                           host_to_target_data_route_rtattr);
 }
 
+static abi_long host_to_target_neigh_rtattr(struct rtattr *rtattr,
+                                         uint32_t rtattr_len)
+{
+    return host_to_target_for_each_rtattr(rtattr, rtattr_len,
+                                          host_to_target_data_neigh_rtattr);
+}
+
 static abi_long host_to_target_data_route(struct nlmsghdr *nlh)
 {
     uint32_t nlmsg_len;
     struct ifinfomsg *ifi;
     struct ifaddrmsg *ifa;
     struct rtmsg *rtm;
+    struct ndmsg *ndm;
 
     nlmsg_len = nlh->nlmsg_len;
     switch (nlh->nlmsg_type) {
@@ -1279,6 +1320,17 @@ static abi_long host_to_target_data_route(struct nlmsghdr *nlh)
                                        nlmsg_len - NLMSG_LENGTH(sizeof(*ifa)));
         }
         break;
+    case RTM_NEWNEIGH:
+    case RTM_DELNEIGH:
+    case RTM_GETNEIGH:
+        if (nlh->nlmsg_len >= NLMSG_LENGTH(sizeof(*ndm))) {
+            ndm = NLMSG_DATA(nlh);
+            ndm->ndm_ifindex = tswap32(ndm->ndm_ifindex);
+            ndm->ndm_state = tswap16(ndm->ndm_state);
+            host_to_target_neigh_rtattr(NDM_RTA(ndm),
+                                    nlmsg_len - NLMSG_LENGTH(sizeof(*ndm)));
+        }
+        break;
     case RTM_NEWROUTE:
     case RTM_DELROUTE:
     case RTM_GETROUTE:
@@ -1426,6 +1478,35 @@ static abi_long target_to_host_data_addr_rtattr(struct rtattr *rtattr)
     return 0;
 }
 
+static abi_long target_to_host_data_neigh_rtattr(struct rtattr *rtattr)
+{
+    struct nda_cacheinfo *ndac;
+    uint32_t *u32;
+
+    switch (rtattr->rta_type) {
+    case NDA_UNSPEC:
+    case NDA_DST:
+    case NDA_LLADDR:
+        break;
+    case NDA_PROBES:
+        u32 = RTA_DATA(rtattr);
+        *u32 = tswap32(*u32);
+        break;
+    case NDA_CACHEINFO:
+        ndac = RTA_DATA(rtattr);
+        ndac->ndm_confirmed = tswap32(ndac->ndm_confirmed);
+        ndac->ndm_used      = tswap32(ndac->ndm_used);
+        ndac->ndm_updated   = tswap32(ndac->ndm_updated);
+        ndac->ndm_refcnt    = tswap32(ndac->ndm_refcnt);
+        break;
+    default:
+        qemu_log_mask(LOG_UNIMP, "Unknown target NEIGH type: %d\n",
+                      rtattr->rta_type);
+        break;
+    }
+    return 0;
+}
+
 static abi_long target_to_host_data_route_rtattr(struct rtattr *rtattr)
 {
     uint32_t *u32;
@@ -1464,6 +1545,13 @@ static void target_to_host_addr_rtattr(struct rtattr *rtattr,
                                    target_to_host_data_addr_rtattr);
 }
 
+static void target_to_host_neigh_rtattr(struct rtattr *rtattr,
+                                     uint32_t rtattr_len)
+{
+    target_to_host_for_each_rtattr(rtattr, rtattr_len,
+                                   target_to_host_data_neigh_rtattr);
+}
+
 static void target_to_host_route_rtattr(struct rtattr *rtattr,
                                      uint32_t rtattr_len)
 {
@@ -1476,6 +1564,7 @@ static abi_long target_to_host_data_route(struct nlmsghdr *nlh)
     struct ifinfomsg *ifi;
     struct ifaddrmsg *ifa;
     struct rtmsg *rtm;
+    struct ndmsg *ndm;
 
     switch (nlh->nlmsg_type) {
     case RTM_NEWLINK:
@@ -1502,6 +1591,17 @@ static abi_long target_to_host_data_route(struct nlmsghdr *nlh)
                                        NLMSG_LENGTH(sizeof(*ifa)));
         }
         break;
+    case RTM_NEWNEIGH:
+    case RTM_DELNEIGH:
+    case RTM_GETNEIGH:
+        if (nlh->nlmsg_len >= NLMSG_LENGTH(sizeof(*ndm))) {
+            ndm = NLMSG_DATA(nlh);
+            ndm->ndm_ifindex = tswap32(ndm->ndm_ifindex);
+            ndm->ndm_state = tswap16(ndm->ndm_state);
+            target_to_host_neigh_rtattr(NDM_RTA(ndm), nlh->nlmsg_len -
+                                       NLMSG_LENGTH(sizeof(*ndm)));
+        }
+        break;
     case RTM_NEWROUTE:
     case RTM_DELROUTE:
     case RTM_GETROUTE:
-- 
2.47.0

[PATCH v4 6/7] linux-user: netlink: Add missing QEMU_IFLA entries

[deller]

From: Helge Deller <deller@gmx.de>

This fixes the following qemu warnings when building debian gupnp package:
 Unknown host QEMU_IFLA type: 61
 Unknown host QEMU_IFLA type: 58
 Unknown host QEMU_IFLA type: 59
 Unknown host QEMU_IFLA type: 60
 Unknown host QEMU_IFLA type: 32820

QEMU_IFLA type 32820 is actually NLA_NESTED | QEMU_IFLA_PROP_LIST (a nested
entry), which is why rta_type needs to be masked with NLA_TYPE_MASK.

Signed-off-by: Helge Deller <deller@gmx.de>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
---
 linux-user/fd-trans.c | 36 +++++++++++++++++++++++++++++++++++-
 1 file changed, 35 insertions(+), 1 deletion(-)

diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
index 59b622aba9..105b1f0713 100644
--- a/linux-user/fd-trans.c
+++ b/linux-user/fd-trans.c
@@ -161,6 +161,14 @@ enum {
     QEMU_IFLA_PROTO_DOWN_REASON,
     QEMU_IFLA_PARENT_DEV_NAME,
     QEMU_IFLA_PARENT_DEV_BUS_NAME,
+    QEMU_IFLA_GRO_MAX_SIZE,
+    QEMU_IFLA_TSO_MAX_SIZE,
+    QEMU_IFLA_TSO_MAX_SEGS,
+    QEMU_IFLA_ALLMULTI,
+    QEMU_IFLA_DEVLINK_PORT,
+    QEMU_IFLA_GSO_IPV4_MAX_SIZE,
+    QEMU_IFLA_GRO_IPV4_MAX_SIZE,
+    QEMU_IFLA_DPLL_PIN,
     QEMU___IFLA_MAX
 };
 
@@ -1002,6 +1010,22 @@ static abi_long host_to_target_data_vfinfo_nlattr(struct nlattr *nlattr,
     return 0;
 }
 
+static abi_long host_to_target_data_prop_nlattr(struct nlattr *nlattr,
+                                                void *context)
+{
+    switch (nlattr->nla_type) {
+    /* string */
+    case QEMU_IFLA_ALT_IFNAME:
+        break;
+    default:
+        qemu_log_mask(LOG_UNIMP, "Unknown host PROP type: %d\n",
+                      nlattr->nla_type);
+        break;
+    }
+    return 0;
+}
+
+
 static abi_long host_to_target_data_link_rtattr(struct rtattr *rtattr)
 {
     uint32_t *u32;
@@ -1010,7 +1034,7 @@ static abi_long host_to_target_data_link_rtattr(struct rtattr *rtattr)
     struct rtnl_link_ifmap *map;
     struct linkinfo_context li_context;
 
-    switch (rtattr->rta_type) {
+    switch (rtattr->rta_type & NLA_TYPE_MASK) {
     /* binary stream */
     case QEMU_IFLA_ADDRESS:
     case QEMU_IFLA_BROADCAST:
@@ -1048,6 +1072,12 @@ static abi_long host_to_target_data_link_rtattr(struct rtattr *rtattr)
     case QEMU_IFLA_CARRIER_DOWN_COUNT:
     case QEMU_IFLA_MIN_MTU:
     case QEMU_IFLA_MAX_MTU:
+    case QEMU_IFLA_GRO_MAX_SIZE:
+    case QEMU_IFLA_TSO_MAX_SIZE:
+    case QEMU_IFLA_TSO_MAX_SEGS:
+    case QEMU_IFLA_ALLMULTI:
+    case QEMU_IFLA_GSO_IPV4_MAX_SIZE:
+    case QEMU_IFLA_GRO_IPV4_MAX_SIZE:
         u32 = RTA_DATA(rtattr);
         *u32 = tswap32(*u32);
         break;
@@ -1143,6 +1173,10 @@ static abi_long host_to_target_data_link_rtattr(struct rtattr *rtattr)
         return host_to_target_for_each_nlattr(RTA_DATA(rtattr), rtattr->rta_len,
                                               NULL,
                                              host_to_target_data_vfinfo_nlattr);
+    case QEMU_IFLA_PROP_LIST:
+        return host_to_target_for_each_nlattr(RTA_DATA(rtattr), rtattr->rta_len,
+                                              NULL,
+                                             host_to_target_data_prop_nlattr);
     default:
         qemu_log_mask(LOG_UNIMP, "Unknown host QEMU_IFLA type: %d\n",
                       rtattr->rta_type);
-- 
2.47.0

[PATCH v4 7/7] linux-user: netlink: Use QEMU_IFA_XXX values instead of IFA_XXX

[deller]

From: Helge Deller <deller@gmx.de>

Convert existing places to use QEMU_IFA_XXX values.

Signed-off-by: Helge Deller <deller@gmx.de>
---
 linux-user/fd-trans.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
index 105b1f0713..f83d1f79d5 100644
--- a/linux-user/fd-trans.c
+++ b/linux-user/fd-trans.c
@@ -1192,21 +1192,21 @@ static abi_long host_to_target_data_addr_rtattr(struct rtattr *rtattr)
 
     switch (rtattr->rta_type) {
     /* binary: depends on family type */
-    case IFA_ADDRESS:
-    case IFA_LOCAL:
+    case QEMU_IFA_ADDRESS:
+    case QEMU_IFA_LOCAL:
     case QEMU_IFA_PROTO:
         break;
     /* string */
-    case IFA_LABEL:
+    case QEMU_IFA_LABEL:
         break;
     /* u32 */
-    case IFA_FLAGS:
-    case IFA_BROADCAST:
+    case QEMU_IFA_FLAGS:
+    case QEMU_IFA_BROADCAST:
         u32 = RTA_DATA(rtattr);
         *u32 = tswap32(*u32);
         break;
     /* struct ifa_cacheinfo */
-    case IFA_CACHEINFO:
+    case QEMU_IFA_CACHEINFO:
         ci = RTA_DATA(rtattr);
         ci->ifa_prefered = tswap32(ci->ifa_prefered);
         ci->ifa_valid = tswap32(ci->ifa_valid);
@@ -1501,8 +1501,8 @@ static abi_long target_to_host_data_addr_rtattr(struct rtattr *rtattr)
 {
     switch (rtattr->rta_type) {
     /* binary: depends on family type */
-    case IFA_LOCAL:
-    case IFA_ADDRESS:
+    case QEMU_IFA_LOCAL:
+    case QEMU_IFA_ADDRESS:
         break;
     default:
         qemu_log_mask(LOG_UNIMP, "Unknown target IFA type: %d\n",
-- 
2.47.0

Re: [PATCH v4 7/7] linux-user: netlink: Use QEMU_IFA_XXX values instead of IFA_XXX

[Laurent Vivier]

Le 20/01/2025 à 22:33, deller@kernel.org a écrit :
> From: Helge Deller <deller@gmx.de>
> 
> Convert existing places to use QEMU_IFA_XXX values.
> 
> Signed-off-by: Helge Deller <deller@gmx.de>
> ---
>   linux-user/fd-trans.c | 16 ++++++++--------
>   1 file changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/linux-user/fd-trans.c b/linux-user/fd-trans.c
> index 105b1f0713..f83d1f79d5 100644
> --- a/linux-user/fd-trans.c
> +++ b/linux-user/fd-trans.c
> @@ -1192,21 +1192,21 @@ static abi_long host_to_target_data_addr_rtattr(struct rtattr *rtattr)
>   
>       switch (rtattr->rta_type) {
>       /* binary: depends on family type */
> -    case IFA_ADDRESS:
> -    case IFA_LOCAL:
> +    case QEMU_IFA_ADDRESS:
> +    case QEMU_IFA_LOCAL:
>       case QEMU_IFA_PROTO:
>           break;
>       /* string */
> -    case IFA_LABEL:
> +    case QEMU_IFA_LABEL:
>           break;
>       /* u32 */
> -    case IFA_FLAGS:
> -    case IFA_BROADCAST:
> +    case QEMU_IFA_FLAGS:
> +    case QEMU_IFA_BROADCAST:
>           u32 = RTA_DATA(rtattr);
>           *u32 = tswap32(*u32);
>           break;
>       /* struct ifa_cacheinfo */
> -    case IFA_CACHEINFO:
> +    case QEMU_IFA_CACHEINFO:
>           ci = RTA_DATA(rtattr);
>           ci->ifa_prefered = tswap32(ci->ifa_prefered);
>           ci->ifa_valid = tswap32(ci->ifa_valid);
> @@ -1501,8 +1501,8 @@ static abi_long target_to_host_data_addr_rtattr(struct rtattr *rtattr)
>   {
>       switch (rtattr->rta_type) {
>       /* binary: depends on family type */
> -    case IFA_LOCAL:
> -    case IFA_ADDRESS:
> +    case QEMU_IFA_LOCAL:
> +    case QEMU_IFA_ADDRESS:
>           break;
>       default:
>           qemu_log_mask(LOG_UNIMP, "Unknown target IFA type: %d\n",

I would put this as patch one and adding the QEMU_IFA_XXX enum in this one.

Otherwise both patches (1 and 7) look good.

Thanks,
Laurent