tjh.dev / kernel
Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux
1
fork

Configure Feed

Select the types of activity you want to include in your feed.

tools headers: Update the linux/gfp_types.h copy with the kernel sources

To pick up the changes in:

f3ec502b6755a3bf ("mm/slab: mark alloc tags empty for sheaves allocated with __GFP_NO_OBJ_EXT")
241b3a09639c317b ("mm: clarify GFP_ATOMIC/GFP_NOWAIT doc-comment")

That just adds some comments, so no changes in perf tooling, just silences
this build warning:

Warning: Kernel ABI header differences:
diff -u tools/include/linux/gfp_types.h include/linux/gfp_types.h

Please see tools/include/uapi/README.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Arnaldo Carvalho de Melo 916a9f38 ecd5a2fd

+7 -2
+7 -2
tools/include/linux/gfp_types.h
··· 139 139 * %__GFP_ACCOUNT causes the allocation to be accounted to kmemcg. 140 140 * 141 141 * %__GFP_NO_OBJ_EXT causes slab allocation to have no object extension. 142 + * mark_obj_codetag_empty() should be called upon freeing for objects allocated 143 + * with this flag to indicate that their NULL tags are expected and normal. 142 144 */ 143 145 #define __GFP_RECLAIMABLE ((__force gfp_t)___GFP_RECLAIMABLE) 144 146 #define __GFP_WRITE ((__force gfp_t)___GFP_WRITE) ··· 311 309 * 312 310 * %GFP_ATOMIC users can not sleep and need the allocation to succeed. A lower 313 311 * watermark is applied to allow access to "atomic reserves". 314 - * The current implementation doesn't support NMI and few other strict 315 - * non-preemptive contexts (e.g. raw_spin_lock). The same applies to %GFP_NOWAIT. 312 + * The current implementation doesn't support NMI, nor contexts that disable 313 + * preemption under PREEMPT_RT. This includes raw_spin_lock() and plain 314 + * preempt_disable() - see "Memory allocation" in 315 + * Documentation/core-api/real-time/differences.rst for more info. 316 316 * 317 317 * %GFP_KERNEL is typical for kernel-internal allocations. The caller requires 318 318 * %ZONE_NORMAL or a lower zone for direct access but can direct reclaim. ··· 325 321 * %GFP_NOWAIT is for kernel allocations that should not stall for direct 326 322 * reclaim, start physical IO or use any filesystem callback. It is very 327 323 * likely to fail to allocate memory, even for very small allocations. 324 + * The same restrictions on calling contexts apply as for %GFP_ATOMIC. 328 325 * 329 326 * %GFP_NOIO will use direct reclaim to discard clean pages or slab pages 330 327 * that do not require the starting of any physical IO.