| Message ID | 20180523232246.20445-1-qing.huang@oracle.com (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
On Thu, May 24, 2018 at 1:22 AM, Qing Huang <qing.huang@oracle.com> wrote: > When a system is under memory presure (high usage with fragments), > the original 256KB ICM chunk allocations will likely trigger kernel > memory management to enter slow path doing memory compact/migration > ops in order to complete high order memory allocations. > > When that happens, user processes calling uverb APIs may get stuck > for more than 120s easily even though there are a lot of free pages > in smaller chunks available in the system. > > Syslog: > ... > Dec 10 09:04:51 slcc03db02 kernel: [397078.572732] INFO: task > oracle_205573_e:205573 blocked for more than 120 seconds. > ... > > With 4KB ICM chunk size on x86_64 arch, the above issue is fixed. > > However in order to support smaller ICM chunk size, we need to fix > another issue in large size kcalloc allocations. > > E.g. > Setting log_num_mtt=30 requires 1G mtt entries. With the 4KB ICM chunk > size, each ICM chunk can only hold 512 mtt entries (8 bytes for each mtt > entry). So we need a 16MB allocation for a table->icm pointer array to > hold 2M pointers which can easily cause kcalloc to fail. > > The solution is to use kvzalloc to replace kcalloc which will fall back > to vmalloc automatically if kmalloc fails. Hi, Could you please write why it first try to allocate the contiguous pages? I think it is necessary to comment why it uses kvzalloc instead of vzalloc. > > Signed-off-by: Qing Huang <qing.huang@oracle.com> > Acked-by: Daniel Jurgens <danielj@mellanox.com> > Reviewed-by: Zhu Yanjun <yanjun.zhu@oracle.com> +Reviewed-by: Gioh Kim <gi-oh.kim@profitbricks.com> > --- > v4: use kvzalloc instead of vzalloc > add one err condition check > don't include vmalloc.h any more > > v3: use PAGE_SIZE instead of PAGE_SHIFT > add comma to the end of enum variables > include vmalloc.h header file to avoid build issues on Sparc > > v2: adjusted chunk size to reflect different architectures > > drivers/net/ethernet/mellanox/mlx4/icm.c | 16 +++++++++------- > 1 file changed, 9 insertions(+), 7 deletions(-) > > diff --git a/drivers/net/ethernet/mellanox/mlx4/icm.c b/drivers/net/ethernet/mellanox/mlx4/icm.c > index a822f7a..685337d 100644 > --- a/drivers/net/ethernet/mellanox/mlx4/icm.c > +++ b/drivers/net/ethernet/mellanox/mlx4/icm.c > @@ -43,12 +43,12 @@ > #include "fw.h" > > /* > - * We allocate in as big chunks as we can, up to a maximum of 256 KB > - * per chunk. > + * We allocate in page size (default 4KB on many archs) chunks to avoid high > + * order memory allocations in fragmented/high usage memory situation. > */ > enum { > - MLX4_ICM_ALLOC_SIZE = 1 << 18, > - MLX4_TABLE_CHUNK_SIZE = 1 << 18 > + MLX4_ICM_ALLOC_SIZE = PAGE_SIZE, > + MLX4_TABLE_CHUNK_SIZE = PAGE_SIZE, > }; > > static void mlx4_free_icm_pages(struct mlx4_dev *dev, struct mlx4_icm_chunk *chunk) > @@ -398,9 +398,11 @@ int mlx4_init_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table, > u64 size; > > obj_per_chunk = MLX4_TABLE_CHUNK_SIZE / obj_size; > + if (WARN_ON(!obj_per_chunk)) > + return -EINVAL; > num_icm = (nobj + obj_per_chunk - 1) / obj_per_chunk; > > - table->icm = kcalloc(num_icm, sizeof(*table->icm), GFP_KERNEL); > + table->icm = kvzalloc(num_icm * sizeof(*table->icm), GFP_KERNEL); > if (!table->icm) > return -ENOMEM; > table->virt = virt; > @@ -446,7 +448,7 @@ int mlx4_init_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table, > mlx4_free_icm(dev, table->icm[i], use_coherent); > } > > - kfree(table->icm); > + kvfree(table->icm); > > return -ENOMEM; > } > @@ -462,5 +464,5 @@ void mlx4_cleanup_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table) > mlx4_free_icm(dev, table->icm[i], table->coherent); > } > > - kfree(table->icm); > + kvfree(table->icm); > } > -- > 2.9.3 >
On 24/05/2018 2:22 AM, Qing Huang wrote: > When a system is under memory presure (high usage with fragments), > the original 256KB ICM chunk allocations will likely trigger kernel > memory management to enter slow path doing memory compact/migration > ops in order to complete high order memory allocations. > > When that happens, user processes calling uverb APIs may get stuck > for more than 120s easily even though there are a lot of free pages > in smaller chunks available in the system. > > Syslog: > ... > Dec 10 09:04:51 slcc03db02 kernel: [397078.572732] INFO: task > oracle_205573_e:205573 blocked for more than 120 seconds. > ... > > With 4KB ICM chunk size on x86_64 arch, the above issue is fixed. > > However in order to support smaller ICM chunk size, we need to fix > another issue in large size kcalloc allocations. > > E.g. > Setting log_num_mtt=30 requires 1G mtt entries. With the 4KB ICM chunk > size, each ICM chunk can only hold 512 mtt entries (8 bytes for each mtt > entry). So we need a 16MB allocation for a table->icm pointer array to > hold 2M pointers which can easily cause kcalloc to fail. > > The solution is to use kvzalloc to replace kcalloc which will fall back > to vmalloc automatically if kmalloc fails. > > Signed-off-by: Qing Huang <qing.huang@oracle.com> > Acked-by: Daniel Jurgens <danielj@mellanox.com> > Reviewed-by: Zhu Yanjun <yanjun.zhu@oracle.com> > --- > v4: use kvzalloc instead of vzalloc > add one err condition check > don't include vmalloc.h any more > > v3: use PAGE_SIZE instead of PAGE_SHIFT > add comma to the end of enum variables > include vmalloc.h header file to avoid build issues on Sparc > > v2: adjusted chunk size to reflect different architectures > > drivers/net/ethernet/mellanox/mlx4/icm.c | 16 +++++++++------- > 1 file changed, 9 insertions(+), 7 deletions(-) > > diff --git a/drivers/net/ethernet/mellanox/mlx4/icm.c b/drivers/net/ethernet/mellanox/mlx4/icm.c > index a822f7a..685337d 100644 > --- a/drivers/net/ethernet/mellanox/mlx4/icm.c > +++ b/drivers/net/ethernet/mellanox/mlx4/icm.c > @@ -43,12 +43,12 @@ > #include "fw.h" > > /* > - * We allocate in as big chunks as we can, up to a maximum of 256 KB > - * per chunk. > + * We allocate in page size (default 4KB on many archs) chunks to avoid high > + * order memory allocations in fragmented/high usage memory situation. > */ > enum { > - MLX4_ICM_ALLOC_SIZE = 1 << 18, > - MLX4_TABLE_CHUNK_SIZE = 1 << 18 > + MLX4_ICM_ALLOC_SIZE = PAGE_SIZE, > + MLX4_TABLE_CHUNK_SIZE = PAGE_SIZE, > }; > > static void mlx4_free_icm_pages(struct mlx4_dev *dev, struct mlx4_icm_chunk *chunk) > @@ -398,9 +398,11 @@ int mlx4_init_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table, > u64 size; > > obj_per_chunk = MLX4_TABLE_CHUNK_SIZE / obj_size; > + if (WARN_ON(!obj_per_chunk)) > + return -EINVAL; > num_icm = (nobj + obj_per_chunk - 1) / obj_per_chunk; > > - table->icm = kcalloc(num_icm, sizeof(*table->icm), GFP_KERNEL); > + table->icm = kvzalloc(num_icm * sizeof(*table->icm), GFP_KERNEL); > if (!table->icm) > return -ENOMEM; > table->virt = virt; > @@ -446,7 +448,7 @@ int mlx4_init_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table, > mlx4_free_icm(dev, table->icm[i], use_coherent); > } > > - kfree(table->icm); > + kvfree(table->icm); > > return -ENOMEM; > } > @@ -462,5 +464,5 @@ void mlx4_cleanup_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table) > mlx4_free_icm(dev, table->icm[i], table->coherent); > } > > - kfree(table->icm); > + kvfree(table->icm); > } > Thanks Qing. Reviewed-by: Tariq Toukan <tariqt@mellanox.com> -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
From: Qing Huang <qing.huang@oracle.com> Date: Wed, 23 May 2018 16:22:46 -0700 > When a system is under memory presure (high usage with fragments), > the original 256KB ICM chunk allocations will likely trigger kernel > memory management to enter slow path doing memory compact/migration > ops in order to complete high order memory allocations. > > When that happens, user processes calling uverb APIs may get stuck > for more than 120s easily even though there are a lot of free pages > in smaller chunks available in the system. > > Syslog: > ... > Dec 10 09:04:51 slcc03db02 kernel: [397078.572732] INFO: task > oracle_205573_e:205573 blocked for more than 120 seconds. > ... > > With 4KB ICM chunk size on x86_64 arch, the above issue is fixed. > > However in order to support smaller ICM chunk size, we need to fix > another issue in large size kcalloc allocations. > > E.g. > Setting log_num_mtt=30 requires 1G mtt entries. With the 4KB ICM chunk > size, each ICM chunk can only hold 512 mtt entries (8 bytes for each mtt > entry). So we need a 16MB allocation for a table->icm pointer array to > hold 2M pointers which can easily cause kcalloc to fail. > > The solution is to use kvzalloc to replace kcalloc which will fall back > to vmalloc automatically if kmalloc fails. > > Signed-off-by: Qing Huang <qing.huang@oracle.com> > Acked-by: Daniel Jurgens <danielj@mellanox.com> > Reviewed-by: Zhu Yanjun <yanjun.zhu@oracle.com> Applied, thanks. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/net/ethernet/mellanox/mlx4/icm.c b/drivers/net/ethernet/mellanox/mlx4/icm.c index a822f7a..685337d 100644 --- a/drivers/net/ethernet/mellanox/mlx4/icm.c +++ b/drivers/net/ethernet/mellanox/mlx4/icm.c @@ -43,12 +43,12 @@ #include "fw.h" /* - * We allocate in as big chunks as we can, up to a maximum of 256 KB - * per chunk. + * We allocate in page size (default 4KB on many archs) chunks to avoid high + * order memory allocations in fragmented/high usage memory situation. */ enum { - MLX4_ICM_ALLOC_SIZE = 1 << 18, - MLX4_TABLE_CHUNK_SIZE = 1 << 18 + MLX4_ICM_ALLOC_SIZE = PAGE_SIZE, + MLX4_TABLE_CHUNK_SIZE = PAGE_SIZE, }; static void mlx4_free_icm_pages(struct mlx4_dev *dev, struct mlx4_icm_chunk *chunk) @@ -398,9 +398,11 @@ int mlx4_init_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table, u64 size; obj_per_chunk = MLX4_TABLE_CHUNK_SIZE / obj_size; + if (WARN_ON(!obj_per_chunk)) + return -EINVAL; num_icm = (nobj + obj_per_chunk - 1) / obj_per_chunk; - table->icm = kcalloc(num_icm, sizeof(*table->icm), GFP_KERNEL); + table->icm = kvzalloc(num_icm * sizeof(*table->icm), GFP_KERNEL); if (!table->icm) return -ENOMEM; table->virt = virt; @@ -446,7 +448,7 @@ int mlx4_init_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table, mlx4_free_icm(dev, table->icm[i], use_coherent); } - kfree(table->icm); + kvfree(table->icm); return -ENOMEM; } @@ -462,5 +464,5 @@ void mlx4_cleanup_icm_table(struct mlx4_dev *dev, struct mlx4_icm_table *table) mlx4_free_icm(dev, table->icm[i], table->coherent); } - kfree(table->icm); + kvfree(table->icm); }