qemu-e2k/softmmu
David Hildenbrand d94e0bc9ef util/mmap-alloc: Support RAM_NORESERVE via MAP_NORESERVE under Linux
Let's support RAM_NORESERVE via MAP_NORESERVE on Linux. The flag has no
effect on most shared mappings - except for hugetlbfs and anonymous memory.

Linux man page:
  "MAP_NORESERVE: Do not reserve swap space for this mapping. When swap
  space is reserved, one has the guarantee that it is possible to modify
  the mapping. When swap space is not reserved one might get SIGSEGV
  upon a write if no physical memory is available. See also the discussion
  of the file /proc/sys/vm/overcommit_memory in proc(5). In kernels before
  2.6, this flag had effect only for private writable mappings."

Note that the "guarantee" part is wrong with memory overcommit in Linux.

Also, in Linux hugetlbfs is treated differently - we configure reservation
of huge pages from the pool, not reservation of swap space (huge pages
cannot be swapped).

The rough behavior is [1]:
a) !Hugetlbfs:

  1) Without MAP_NORESERVE *or* with memory overcommit under Linux
     disabled ("/proc/sys/vm/overcommit_memory == 2"), the following
     accounting/reservation happens:
      For a file backed map
       SHARED or READ-only - 0 cost (the file is the map not swap)
       PRIVATE WRITABLE - size of mapping per instance

      For an anonymous or /dev/zero map
       SHARED   - size of mapping
       PRIVATE READ-only - 0 cost (but of little use)
       PRIVATE WRITABLE - size of mapping per instance

  2) With MAP_NORESERVE, no accounting/reservation happens.

b) Hugetlbfs:

  1) Without MAP_NORESERVE, huge pages are reserved.

  2) With MAP_NORESERVE, no huge pages are reserved.

Note: With "/proc/sys/vm/overcommit_memory == 0", we were already able
to configure it for !hugetlbfs globally; this toggle now allows
configuring it more fine-grained, not for the whole system.

The target use case is virtio-mem, which dynamically exposes memory
inside a large, sparse memory area to the VM.

[1] https://www.kernel.org/doc/Documentation/vm/overcommit-accounting

Reviewed-by: Peter Xu <peterx@redhat.com>
Acked-by: Eduardo Habkost <ehabkost@redhat.com> for memory backend and machine core
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210510114328.21835-10-david@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2021-06-15 20:27:38 +02:00
..
arch_init.c Drop the deprecated unicore32 target 2021-05-12 18:20:52 +02:00
balloon.c qapi: Restrict balloon-related commands to machine code 2020-09-29 15:41:35 +02:00
bootdevice.c softmmu: move more files to softmmu/ 2020-10-12 11:50:21 -04:00
cpu-throttle.c cpu-throttle: Remove timer_mod() from cpu_throttle_set() 2021-02-08 15:15:32 +01:00
cpu-timers.c icount: get rid of static variable 2021-04-01 09:40:45 +02:00
cpus.c softmmu: add trace point when bdrv_flush_all fails 2021-06-14 13:28:50 +01:00
datadir.c vl: extract softmmu/datadir.c 2020-12-10 12:15:18 -05:00
device_tree.c Do not include sysemu/sysemu.h if it's not really necessary 2021-05-02 17:24:50 +02:00
dma-helpers.c dma: Introduce dma_aligned_pow2_mask() 2021-03-12 12:40:10 +00:00
globals.c qmp: generalize watchdog-set-action to -no-reboot/-no-shutdown 2020-12-15 12:51:57 -05:00
icount.c icount: get rid of static variable 2021-04-01 09:40:45 +02:00
ioport.c softmmu: Add missing trace-events file 2020-09-09 17:15:18 +01:00
main.c meson: move SDL and SDL-image detection to meson 2020-08-21 06:30:44 -04:00
memory_mapping.c Do not include cpu.h if it's not really necessary 2021-05-02 17:24:51 +02:00
memory.c softmmu/memory: Pass ram_flags to qemu_ram_alloc() and qemu_ram_alloc_internal() 2021-06-15 20:27:38 +02:00
meson.build seccomp: convert to meson 2021-01-06 10:21:20 +01:00
physmem.c util/mmap-alloc: Support RAM_NORESERVE via MAP_NORESERVE under Linux 2021-06-15 20:27:38 +02:00
qdev-monitor.c Do not include sysemu/sysemu.h if it's not really necessary 2021-05-02 17:24:50 +02:00
qemu-seccomp.c seccomp: Replace the word 'blacklist' 2021-03-09 22:15:19 +01:00
qtest.c qtest: add a QOM object for qtest 2021-05-26 14:49:45 +02:00
rtc.c qom: Allow optional sugar props 2021-02-08 16:57:37 +11:00
runstate-action.c runstate: cleanup reboot and panic actions 2021-01-21 13:00:41 +01:00
runstate.c runstate: Initialize Error * to NULL 2021-06-15 17:17:09 +02:00
timers-state.h icount: get rid of static variable 2021-04-01 09:40:45 +02:00
tpm.c qapi: More complex uses of QAPI_LIST_APPEND 2021-01-28 08:08:45 +01:00
trace-events softmmu: add trace point when bdrv_flush_all fails 2021-06-14 13:28:50 +01:00
trace.h softmmu: Add missing trace-events file 2020-09-09 17:15:18 +01:00
vl.c vl: Fix an assert failure in error path 2021-06-15 17:17:09 +02:00