Je regarde tout à l'heure comment c'est géré sur le mien.

C'est un serveur avec un bi processeur Intel® Xeon E5-2620 v2.

Chaque processeur à 6 cœurs et 12 threads.

L'organisation, c'est :
- en 1er les 6 cœurs physique du 1er CPU
- en second les 6 cœurs logiques du 1er CPU
- en 3ème, les 6 cœurs physique du 2ème CPU
- en 4ème, les 6 cœurs logiques du 2ème CPU

Les "cœurs" partagent quoi exactement?

Cœurs qui existent réellement.

cœurs logiques = cœurs virtuels = cœurs bidons

Sous Linux:

Each CPU has a "base" scheduling domain (struct sched_domain). The domain
hierarchy is built from these base domains via the ->parent pointer. ->parent
MUST be NULL terminated, and domain structures should be per-CPU as they are
locklessly updated.

Each scheduling domain spans a number of CPUs (stored in the ->span field).
A domain's span MUST be a superset of it child's span (this restriction could
be relaxed if the need arises), and a base domain for CPU i MUST span at least
i. The top domain for each CPU will generally span all CPUs in the system
although strictly it doesn't have to, but this could lead to a case where some
CPUs will never be given tasks to run unless the CPUs allowed mask is
explicitly set. A sched domain's span means "balance process load among these
CPUs".

Each scheduling domain must have one or more CPU groups (struct sched_group)
which are organised as a circular one way linked list from the ->groups
pointer. The union of cpumasks of these groups MUST be the same as the
domain's span. The intersection of cpumasks from any two of these groups
MUST be the empty set. The group pointed to by the ->groups pointer MUST
contain the CPU to which the domain belongs. Groups may be shared among
CPUs as they contain read only data after they have been set up.

Balancing within a sched domain occurs between groups. That is, each group
is treated as one entity. The load of a group is defined as the sum of the
load of each of its member CPUs, and only when the load of a group becomes
out of balance are tasks moved between groups.

In kernel/sched/core.c, trigger_load_balance() is run periodically on each CPU
through scheduler_tick(). It raises a softirq after the next regularly scheduled
rebalancing event for the current runqueue has arrived. The actual load
balancing workhorse, run_rebalance_domains()->rebalance_domains(), is then run
in softirq context (SCHED_SOFTIRQ).

The latter function takes two arguments: the current CPU and whether it was idle
at the time the scheduler_tick() happened and iterates over all sched domains
our CPU is on, starting from its base domain and going up the ->parent chain.
While doing that, it checks to see if the current domain has exhausted its
rebalance interval. If so, it runs load_balance() on that domain. It then checks
the parent sched_domain (if it exists), and the parent of the parent and so
forth.

Initially, load_balance() finds the busiest group in the current sched domain.
If it succeeds, it looks for the busiest runqueue of all the CPUs' runqueues in
that group. If it manages to find such a runqueue, it locks both our initial
CPU's runqueue and the newly found busiest one and starts moving tasks from it
to our runqueue. The exact number of tasks amounts to an imbalance previously
computed while iterating over this sched domain's groups.

*** Implementing sched domains ***
The "base" domain will "span" the first level of the hierarchy. In the case
of SMT, you'll span all siblings of the physical CPU, with each group being
a single virtual CPU.

In SMP, the parent of the base domain will span all physical CPUs in the
node. Each group being a single physical CPU. Then with NUMA, the parent
of the SMP domain will span the entire machine, with each group having the
cpumask of a node. Or, you could do multi-level NUMA or Opteron, for example,
might have just one domain covering its one NUMA level.

The implementor should read comments in include/linux/sched.h:
struct sched_domain fields, SD_FLAG_*, SD_*_INIT to get an idea of
the specifics and what to tune.

Architectures may retain the regular override the default SD_*_INIT flags
while using the generic domain builder in kernel/sched/core.c if they wish to
retain the traditional SMT->SMP->NUMA topology (or some subset of that). This
can be done by #define'ing ARCH_HASH_SCHED_TUNE.

Alternatively, the architecture may completely override the generic domain
builder by #define'ing ARCH_HASH_SCHED_DOMAIN, and exporting your
arch_init_sched_domains function. This function will attach domains to all
CPUs using cpu_attach_domain.

The sched-domains debugging infrastructure can be enabled by enabling
CONFIG_SCHED_DEBUG. This enables an error checking parse of the sched domains
which should catch most possible errors (described above). It also prints out
the domain structure in a visual format.

Pour résumer, les CPU sont répartis dans des struct sched_domain.
Par défaut, sur les systèmes multi-sockets, chaque socket est dans un domaine séparé.
Sur les systèmes multi-threads (le cas de l'hyperthreading, donc), chaque groupe de core (le physique et le ou les virtuel(s) sont dans un domaine séparé.

Lorsqu'une tâche demande de la ressource, elle est alloué à un groupe (le plus vide, j'imagine).
Lorsqu'un groupe est "saturé", des tâches peuvent être exportés vers un autre groupe.

Cela signifie qu'un process reste toujours dans le même groupe (sauf saturation CPU) et profite donc grandement du cache.
Cela signifie également que deux process séparés font être alloués à deux cores physiques si possible.