When the mempool uses the waitsem semaphore for memory allocation,
the flowing two conditions must be satisfied simultaneously:
its wait variable is set to true, and the expandsize variable is set to 0
Signed-off-by: zhanxiaoqi <zhanxiaoqi@bytedance.com>
The wait variable of the memory pool is modified to be controlled by macros, facilitating dynamic adjustment of its value via configuration macros.
Signed-off-by: zhanxiaoqi <zhanxiaoqi@bytedance.com>
The wait member variable of the memory pool structure is not initialized, leading to undefined behavior of the memory pool.
Signed-off-by: zhanxiaoqi <zhanxiaoqi@bytedance.com>
Intermediate files of make depend like .ddc and .dds may remain
when make is interrupted. Remove them using make distclean.
Signed-off-by: SPRESENSE <41312067+SPRESENSE@users.noreply.github.com>
Use size_t instead of unsigned int for 'mask' to avoid address truncation,
especially when memory addresses exceed 32 bits.
Signed-off-by: liang.huang <liang.huang@houmo.ai>
and migrate arch/sim from the customized mm_heap to
umm_heap, so the default mm_heap implementation can
still be used(e.g. shared memory in OpenAMP).
Signed-off-by: ganjing <ganjing@xiaomi.com>
Stop the report handler from re-entering KASAN by halting checks when printing.
Reuse the dump_only flag so read/write panic toggles only emit stack traces.
Retain stack dumps for panic-disabled paths without risking another overflow.
Signed-off-by: anpeiyun <anpeiyun@xiaomi.com>
Support enable/disable the kasan when initialize the heap.
This requirement is from rptun, because rptun use share memory init
the heap and share memory is precious, so we need disable the kasan
feature for rptun's heap to save the share memory.
Signed-off-by: Bowen Wang <wangbowen6@xiaomi.com>
Signed-off-by: wangzhi16 <wangzhi16@xiaomi.com>
Support user pass it own heap struct to the mm_initialize_heap() to
avoid the heap struct is reserved from the heap range
Signed-off-by: anjiahao <anjiahao@xiaomi.com>
Update io_len to 0 when pktlen is 0. otherwise, io_pktlen and io_len will be inconsistent, cause free check failure.
Signed-off-by: gaohedong <gaohedong@xiaomi.com>
This interface allows different protocol modules to use their own unique IOB
buffer sources and allocation strategies without interfering with each other.
Representative examples are the IP protocol and the CAN protocol. The IP
protocol generally transmits packets of varying lengths and requires
relatively high peak throughput. In this case, to ensure higher performance,
IOB_BUFSIZE is often configured to be relatively large, such as greater than
500. The CAN protocol generally transmits short packets of fixed length.
In this case, to improve memory utilization, IOB_BUFSIZE is often configured
to be relatively small, such as 16 or 64. To optimize the memory utilization
when the IP protocol and the CAN protocol share the same core,
this interface was added.
Signed-off-by: zhanghongyu <zhanghongyu@xiaomi.com>
We don't need to subtract the block size; we only need to determine if it's within the interrupt memory range.
Signed-off-by: anjiahao <anjiahao@xiaomi.com>
remove reference to non-existent readme in mm/
Pointing to the documentation page doesn't make sense in this case,
because it doesn't explain the use of `#undef XXX` for this case anyway.
Signed-off-by: raiden00pl <raiden00@railab.me>
When initializing a memory block, the shadow area record of the first
memory block is used first.When uninitializing, unpoison is required, otherwise the memory will be marked incorrectly.
The following case will cause problems:
void *mem = malloc(1024);
struct mm_heap_s *a = mm_initialize("hello", mem, 1024);
int *b = mm_malloc(a, sizeof(int *));
*b = 100;
printf("Hello, World!! %d\n", *b);
mm_free(a, b);
mm_uninitialize(a);
free(mem);
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
This can avoid crashes caused by uninitialized accesses to initialized variables in non-chip memory.
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
1. Add hw_tags.c, which will call arm64_mte to implement tagging of memory blocks by operating registers
2. It has been able to run normally on the default NX memory allocator, excluding mempool and tlsf
3. On more complex configurations, memory tests such as memstress can run normally without system crashes
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
When the size of the new realloc is larger than the old one and can be expanded forward and backward, the tag of oldmem needs to be set to the same as newmem, otherwise memcpy will report a kasan error.
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
After it is not zero, the preceding member of the next node will no longer belong to the valid area of the previous alloc node.
Due to the existence of precedence, the memory block size of the node can only be aligned with sizeof(mmsize_t).
This configuration will be applied in the following scenarios when set 8:
ARM64 MTE hardware tag KASan, which requires the tag's memory address to be 16-byte aligned and the memory size must also be 16-byte aligned
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
Renaming "modlib" to "libelf" is more in line with the implementation content,
which makes it easier for individual developers to understand the capabilities of this module.
CONFIG_LIBC_MODLIB -> CONFIG_LIBC_ELF
Signed-off-by: chao an <anchao.archer@bytedance.com>
Fix a memory corruption & deadlocks while calling gran_free.
The function expects a GRAN_HANDLE as a paramter, not a pointer to it.
Signed-off-by: Jukka Laitinen <jukkax@ssrc.tii.ae>
It will be used to distinguish between hardware KASan and software KASan. Hardware KASan does not need to use plug-in
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
If thread 1 is executing kasan_unpoison but a scheduling occurs and the block is trampled upon, the displayed backtracking may still be from the previously allocated backtracking
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
