Instruction
WRFSBASE
Writes the FS base from a register, so a thread can change its TLS pointer without a system call; 64-bit mode, with CR4.FSGSBASE.
- Extensions
- Forms
- 1 (1 Legacy)
- Data
- Intel XED v2026.08.23
WRFSBASE loads the base address of the FS segment from a general
register, so that later FS-relative accesses add the new base. It is
the user-mode counterpart of arch_prctl(ARCH_SET_FS, ...), without
the system call: one instruction moves a thread’s
thread-local storage to another block. It changes no flag
and leaves the FS selector alone. RDFSBASE reads the
base back, and WRGSBASE writes GS the same way. Like
RDFSBASE, it raises #UD unless the processor has FSGSBASE and the
kernel has set CR4.FSGSBASE, which Linux reports as HWCAP2_FSGSBASE
in the auxiliary vector.
Canonical bases only
The value must be a canonical address, or WRFSBASE raises #GP(0) and the base keeps its old value: Zen 5 did so for 8000000000000000H, and accepted FFFF800012345000H and 00007FFFFFFFF000H. The check is on the value written, not on any access made later.
The 32-bit form
Without REX.W, the source is a 32-bit register. Intel’s manual clears bits 63:32 of the base; AMD’s says only that the register goes to the lower 32 bits of FS.base. Zen 5 cleared them: with ECX = 80001234H under random upper bits of RCX, the base became 0000000080001234H, not a merge with the old one.
FS belongs to the C library
Linux’s documentation describes FS as the thread pointer that runtimes
and threading libraries manage, and asks programs that use such
libraries not to repurpose it; GS has no common use in user space and
can serve the program instead. A program that writes FS under glibc
moves every __thread variable and the stack protector’s canary along
with it. Linux promises to preserve the full base of each thread
along with its selector, so a base written this way stays with the
thread.
GCC 14 and Clang 23 expose it as _writefsbase_u64() and
_writefsbase_u32() with -mfsgsbase; at -O2 a call is a single
wrfsbase rdi. AMD’s Zen 5 guide lists WRFSBASE as microcoded.
Forms
| Instruction | Encoding | Requires |
|---|---|---|
WRFSBASE r32/64 64-bit mode only
WRFSBASE_GPRy | F3 0F AE /2 | FSGSBASE |
Operation
What the forms do, in pseudocode that our tests run against results recorded on real processors. How to read.
v = src.read() # r32: zero-extended
if not canonical(v):
fault(GP, 0)
fsbase = vRun on AMD Zen 5 (EPYC 9645): 4 cases.
Operands
| Operand | Kind | Action |
|---|---|---|
WRFSBASE r32/64 64-bit mode only | ||
r32/64 | explicit | read |
FS base | implicit | write |
Flags
None: XED records no flag that the forms test or change.
Attributes
What else XED records about the forms. How to read.
- All forms
- may abort a TSX transaction
Intrinsics
The C intrinsics of Clang 23.1.2 that compile to WRFSBASE, each with the form it compiles to and the Clang options it needs. The forms come from compiling every intrinsic and decoding the result with Intel XED; all intrinsics A–Z explains the method.
| Intrinsic | Compiles to |
|---|---|
_writefsbase_u32
void _writefsbase_u32(unsigned int __V) |
|
_writefsbase_u64
void _writefsbase_u64(unsigned long long __V) |
|
CPUID and processors
Each form belongs to an XED ISA set. A form can be used when the processor reports every CPUID bit of one of its ISA set's alternatives.
| ISA set | CPUID | Processors |
|---|---|---|
RDWRFSGS | FSGSBASE | Ivy Bridge, Haswell, Broadwell, Skylake, Comet Lake, Cannon Lake, Ice Lake (client), Tiger Lake, Alder Lake, Arrow Lake, Panther Lake, Nova Lake, Skylake-SP, Cascade Lake, Cooper Lake, Ice Lake (server), Sapphire Rapids, Emerald Rapids, Granite Rapids, Diamond Rapids, Goldmont, Goldmont Plus, Tremont, Snow Ridge, Lakefield, Sierra Forest, Clearwater Forest, Knights Landing, Knights Mill, Bulldozer, Piledriver, Zen, Zen+, Zen 2, Zen 3, Zen 4, Zen 5, Zen 6 |
More FSGSBASE instructions
Sources
- Intel XED v2026.08.23 (commit
0bcb6237345c): forms, encodings, operands, flags, exception classes, attributes, ISA sets, CPUID bits and chip model. - The CPU data of x86-64.net: Zen 3 to Zen 6, from GCC’s
-marchfeature sets and AMD’s Processor Programming References, which the page of each processor cites. - Clang's x86 intrinsic headers, LLVM 23.1.2 (commit
85ac56026243): the intrinsics, their signatures and doc comments, and the Clang of that release, which compiled each of them. - Intel 64 and IA-32 Architectures Software Developer's Manual, Volume 2 (325383-093, September 2026): Volume 2D, page 6-6 (WRFSBASE/
WRGSBASE) - Linux kernel documentation: Using FS and GS segments in user space applications
- AMD64 Architecture Programmer's Manual, Volume 3: General-Purpose and System Instructions (24594, rev. 3.36, March 2024), Wayback Machine copy of AMD's PDF
- Software Optimization Guide for the AMD Zen5 Microarchitecture (58455, rev. 1.00, August 2024), Wayback Machine copy of AMD's ZIP
The tables are derived from Intel XED, Copyright Intel Corporation, licensed under the Apache License 2.0; x86-64.net converted and reformatted them. The intrinsics and their signatures come from Clang, part of the LLVM Project, licensed under the Apache License 2.0 with LLVM Exceptions. The text is our own.