Re: [PATCH] eal/x86: fix 32-bit write-combined stores

2024-10-04 Thread David Marchand
On Fri, Sep 6, 2024 at 3:28 PM Bruce Richardson wrote: > > The "movdiri" instruction is given as a series of bytes in rte_io.h so > that it works on compilers/assemblers which are unaware of the > instruction. The REX prefix (0x40) on this instruction is invalid for > 32-bit code, causing issues.

Re: [PATCH] eal/x86: fix 32-bit write-combined stores

2024-09-11 Thread Tyler Retzlaff
On Fri, Sep 06, 2024 at 02:27:57PM +0100, Bruce Richardson wrote: > The "movdiri" instruction is given as a series of bytes in rte_io.h so > that it works on compilers/assemblers which are unaware of the > instruction. The REX prefix (0x40) on this instruction is invalid for > 32-bit code, causing

Re: [PATCH] eal/x86: fix 32-bit write-combined stores

2024-09-06 Thread Radu Nicolau
On 06-Sep-24 2:27 PM, Bruce Richardson wrote: The "movdiri" instruction is given as a series of bytes in rte_io.h so that it works on compilers/assemblers which are unaware of the instruction. The REX prefix (0x40) on this instruction is invalid for 32-bit code, causing issues. Thankfully, the

[PATCH] eal/x86: fix 32-bit write-combined stores

2024-09-06 Thread Bruce Richardson
The "movdiri" instruction is given as a series of bytes in rte_io.h so that it works on compilers/assemblers which are unaware of the instruction. The REX prefix (0x40) on this instruction is invalid for 32-bit code, causing issues. Thankfully, the prefix is unnecessary in 64-bit code, since the da