| From | David Laight <> | Subject | RE: [PATCH 0/4] -ffreestanding/-fno-builtin-* patches | Date | Wed, 19 Aug 2020 08:26:51 +0000 |
| |
From: Arvind Sankar > Sent: 18 August 2020 23:26 ... > gcc-10 optimizes the generic memset implementation in there into a call > to memset. Now that's on x86 which doesn't use the generic > implementation, but this is just waiting to bite us. > > https://godbolt.org/z/6EhG15
Gah, if I want a call to memset() I'll write a call to memset(). If I'm trying to zero a small number of bytes I'll write a loop because it'll be faster.
I wish compilers wouldn't even consider these sorts of 'optimisations'.
David
- Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
|