llama: update to commit 71e90e88 (#10192)

llama/llama.cpp/src/llama-mmap.cpp

@@ -8,6 +8,7 @@
 #include <climits>
 #include <stdexcept>
 #include <cerrno>
+#include <algorithm>
 
 #ifdef __has_include
     #if __has_include(<unistd.h>)
@@ -34,6 +35,10 @@
     #include <io.h>
 #endif
 
+#if defined(__APPLE__)
+#include <TargetConditionals.h>
+#endif
+
 // TODO: consider moving to llama-impl.h if needed in more places
 #if defined(_WIN32)
 static std::string llama_format_win_err(DWORD err) {
@@ -471,7 +476,11 @@ struct llama_mlock::impl {
 
         char* errmsg = std::strerror(errno);
         bool suggest = (errno == ENOMEM);
-
+#if defined(TARGET_OS_VISION) || defined(TARGET_OS_TV) || defined(_AIX)
+        // visionOS/tvOS dont't support RLIMIT_MEMLOCK
+        // Skip resource limit checks on visionOS/tvOS
+        suggest = false;
+#else
         struct rlimit lock_limit;
         if (suggest && getrlimit(RLIMIT_MEMLOCK, &lock_limit)) {
             suggest = false;
@@ -479,6 +488,7 @@ struct llama_mlock::impl {
         if (suggest && (lock_limit.rlim_max > lock_limit.rlim_cur + size)) {
             suggest = false;
         }
+#endif
 
         LLAMA_LOG_WARN("warning: failed to mlock %zu-byte buffer (after previously locking %zu bytes): %s\n%s",
                 size, this->size, errmsg, suggest ? MLOCK_SUGGESTION : "");