Ping: https://gcc.gnu.org/ml/gcc-patches/2018-11/msg01674.html
On 11/19/18 6:14 PM, Martin Sebor wrote:
The recent -Wbuiltin-declaration-mismatch enhancement to detect calls with incompatible arguments to built-ins declared without a prototype introduced a subtle bug in the detection of floating conversion errors when passing EXCESS_PRECISION_EXPR arguments to floating function parameters with different precision. The bug was caught by tests on a handful of targets, but unfortunately not on x86_64-linux. The attached patch corrects this bug. Because it also changes the indentation of the affected function to make the control flow easier to follow I include two diffs: one that can be applied as is and another with whitespace changes ignored. Tested on x86_64-linux with -m32/-m64 and with an i386-solaris2.11 cross-compiler. Martin
