Lines Matching +full:high +full:- +full:performance
1 # SPDX-License-Identifier: GPL-2.0-only
12 flaws, this plugin is available to identify and zero-initialize
23 def_bool $(cc-option,-ftrivial-auto-var-init=pattern)
26 def_bool $(cc-option,-ftrivial-auto-var-init=zero)
29 # Clang 16 and later warn about using the -enable flag, but it
31 def_bool $(cc-option,-ftrivial-auto-var-init=zero -enable-trivial-auto-var-init-zero-knowing-it-will-be-removed-from-clang)
47 variables initialized), but the performance impact depends
64 bool "zero-init structs marked for userspace (weak)"
69 Zero-initialize any structures on the stack containing
72 exposures, like CVE-2013-2141:
76 bool "zero-init structs passed by reference (strong)"
82 Zero-initialize any structures on the stack that may
86 exposures, like CVE-2017-1000410:
89 As a side-effect, this keeps a lot of variables on the
95 bool "zero-init everything passed by reference (very strong)"
101 Zero-initialize any stack variables that may be passed
107 As a side-effect, this keeps a lot of variables on the
113 bool "pattern-init everything (strongest)"
125 non-NULL values, buffer sizes and indices are very big. The
126 pattern is situation-specific; Clang on 64-bit uses 0xAA
128 which use 0xFF repeating (-NaN). Clang on 32-bit uses 0xFF
133 bool "zero-init everything (strongest and safest)"
144 (immediately NUL-terminated), pointers (NULL), indices
173 most uninitialized stack variable attacks, with the performance
177 The performance impact on a single CPU system kernel compilation
217 can be useful for estimating the STACKLEAK performance impact for
237 heap content exposures. The performance impact varies by
239 workloads have measured as high as 7%.
254 The performance impact varies by workload, but is more expensive
256 touching "cold" memory areas. Most cases see 3-5% impact. Some
257 synthetic workloads have measured as high as 8%.
260 def_bool $(cc-option,-fzero-call-used-regs=used-gpr)
262 # https://github.com/llvm/llvm-project/issues/59242
269 At the end of functions, always zero any caller-used register
275 generated "write-what-where" gadgets) in the resulting kernel
276 image. This has a less than 1% performance impact on most
288 # https://github.com/llvm/llvm-project/issues/53645
321 Minimal integrity checking in the linked-list manipulation routines
340 def_bool $(cc-option,-frandomize-layout-seed-file=/dev/null)
342 # Clang 16 due to https://github.com/llvm/llvm-project/issues/60349
353 marked with __randomize_layout, will be randomized at compile-time.
358 Enabling this feature will introduce some performance impact,
381 memory size and performance impact.
391 bool "Limit randomization of structure layout to cache-lines"
396 best effort at restricting randomization to cacheline-sized
398 in structures. This reduces the performance hit of RANDSTRUCT