doc: fix cargo doc warnings

Two bare links and two broken intra-doc links.

Signed-off-by: Patrick Roy <[email protected]>

Author: roypat

kvm-bindings/src/x86_64/fam_wrappers.rs

@@ -192,7 +192,7 @@ unsafe impl FamStruct for kvm_xsave2 {
 /// cannot happen in the [`FamStruct::len`] trait method. To work around this, we define a wrapper
 /// struct that caches the length of a previous `KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2)` call,
 /// and implement [`FamStruct`] for this wrapper. Then in kvm-ioctls, we can expose a function
-/// that first queries `KVM_CAP_XSAVE2`, then invokes [`KVM_GET_XSAVE2`] to retrieves the
+/// that first queries `KVM_CAP_XSAVE2`, then invokes `KVM_GET_XSAVE2` to retrieves the
 /// `kvm_xsave` structure, and finally combine them into the [`kvm_xsave2`] helper structure to be
 /// managed as a `FamStruct`.
 pub type Xsave = FamStructWrapper<kvm_xsave2>;

kvm-ioctls/src/ioctls/device.rs

@@ -45,9 +45,9 @@ impl DeviceFd {
     ///
     /// Configuring a VFIO device using `set_device_attr`. Note that VFIO
     /// devices are not yet available on RISC-V The patch for QEMU:
-    /// https://lore.kernel.org/all/[email protected]/
+    /// <https://lore.kernel.org/all/[email protected]/>
     /// and patch for linux kernel
-    /// https://github.com/ventanamicro/linux/tree/dev-upstream are both not
+    /// <https://github.com/ventanamicro/linux/tree/dev-upstream> are both not
     /// upstreamed. Disabling VFIO device test for RISC-V at the time being.
     ///
     /// ```rust

kvm-ioctls/src/ioctls/vcpu.rs

@@ -1966,8 +1966,6 @@ impl VcpuFd {
     ///     // Next, serialize the actual state into a file or so.
     /// }
     /// ```
-    ///
-    /// [`Kvm::check_extension_int`]: kvm_ioctls::Kvm::check_extension_int
     #[cfg(target_arch = "x86_64")]
     pub fn nested_state(
         &self,