In SMP mode, if all cores start at same time, all from __start(),
then only primary need do initialize, so others core should wait
primary, use 'sev' let the non-primary continue to __cpuN_start().
Signed-off-by: ligd <liguiding1@xiaomi.com>
Summary:
- I noticed that the OpenSBI library depends on the BSD license
- This commit fixes this issue
Impact:
- None
Testing:
- Build only
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Summary:
- I noticed that the OpenSBI library depends on the BSD license
- This commit fixes this issue
Impact:
- CONFIG_OPENSBI=y only
Testing:
- Build with icicle:opensbi (will be updated later)
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Optimize sched_note_begin/end, replace note_printf with note_string
sched_note_begin/end optimized from 50us to 1us per consumption
Signed-off-by: yinshengkai <yinshengkai@xiaomi.com>
NON-primary cpu will invalidate cpu0's cache L2, that will caused cpu0's data mismatch, and then system crash
Signed-off-by: ligd <liguiding1@xiaomi.com>
The IDMA needs to have 32 bit word alignment, in fact it will
AND off the lower 2 bits of the value stored in IDMABASE0R.
This bug was masked by CONFIG_ARMV7M_DCACHE causing proper word alignment
and also FAT_DMAMEMORY being aligned.
This commit extends the unaligned logic (used for dcache) to take into account
the need for a buffer copy when the buffer is ot 32 bit word.
It leverages the fact that when CONFIG_ARMV7M_DCACHE is not defined the up_xxxxx_dcache are nops.
1. Don't handle invalid or empty pte entries
num_partition_entries field in GPT typically means number of maximum entries
and not the number of used entries. Empty entries are indentified with
"0" partition type guid. Loop through all the entries
2. Fix the GPT partition size calculation
"ending_lba" is included in the partition, it is not the start of the next one.
Thus the correct size of the partition is end-start+1
Signed-off-by: Jukka Laitinen <jukkax@ssrc.tii.ae>
NuttShell (NSH) NuttX-10.3.0
ap> ifconfig
=================================================================
Program received signal SIGSEGV, Segmentation fault.
==3920365==ERROR: AddressSanitizer: global-buffer-overflow on address 0x57fb4f2a at pc 0x57177067 bp 0xf1ffebb8 sp 0xf1ffeba8
READ of size 1 at 0x57fb4f2a thread T0
...................
| #10 0xf7ac4339 in __asan::__asan_report_load1 (addr=1476087594) at ../../../../../src/libsanitizer/asan/asan_rtl.cpp:117
| #11 0x57177067 in strncmp (cs=0x57fb4f2a "", ct=0x582d36e0 "stat", nb=4) at string/lib_strncmp.c:40
| #12 0x57f3b467 in netprocfs_opendir (relpath=0x57fb4f26 "net", dir=0xf1ffed80) at procfs/net_procfs.c:398
| #13 0x572b3ae1 in procfs_opendir (mountpt=0xf4602c20, relpath=0x57fb4f26 "net", dir=0xf1ffed80) at procfs/fs_procfs.c:625
| #14 0x572879ff in open_mountpoint (inode=0xf4602c20, relpath=0x57fb4f26 "net", dir=0xf1ffed80) at vfs/fs_dir.c:127
...................
Signed-off-by: chao an <anchao@xiaomi.com>
Fix build break:
Error: L6218E: Undefined symbol _sbss (referred from arm_head.o).
Error: L6218E: Undefined symbol _ebss (referred from arm_head.o).
Error: L6218E: Undefined symbol _eronly (referred from arm_head.o).
Error: L6218E: Undefined symbol _sdata (referred from arm_head.o).
Error: L6218E: Undefined symbol _edata (referred from arm_head.o).
Signed-off-by: chao an <anchao@xiaomi.com>
* boards/arm/tiva/tm4c129e-launchpad/README.txt:
Expand this documentation with information about: toolchains,
debugging (including the elusive openocd incantation to flash program
the board), MCU clocking, more details about the Virtual COM port
exposed via ICDI, and a description of the recently added ostest
configuration.
* boards/arm/stm32/b-g474e-dpow1/README.txt:
(Debugging): OpenOCD requires permissions for certain operations. The
preferred thing to do is to properly configure udev rules. The
expedient thing to do is to run it as root with 'sudo', but we do not
want to encourage that, as discussed earlier on this PR [1]. This
revised text offers a compromise where we mention both options, but
we talk about udev rules first and explicitly say that using 'sudo'
is not encouraged. We leave it up to developers to decide what is
best in their particular situation.
References:
[1] https://github.com/apache/incubator-nuttx/pull/7177
Summary:
- I noticed that lc823450-xgevk does not boot due to the recent
changes on g_current_regs
- This PR fixes this issue
Impact:
- None
Testing:
- Tested with lc823450-xgevk:rndis
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
* boards/arm/stm32/b-g474e-dpow1/configs/ostest/defconfig:
New file. Compared to b-g474e-dpow1:nsh, it adds the following
configs:
+CONFIG_BUILTIN=y
+CONFIG_NSH_BUILTIN_APPS=y
+CONFIG_TESTING_OSTEST=y
and, because ostest currently fails when priority inheritance is
enabled, it removes the following configs:
-CONFIG_PRIORITY_INHERITANCE=y
-CONFIG_PTHREAD_MUTEX_DEFAULT_PRIO_INHERIT=y
1. some arm instructions are not compatible with arch tlsr:
{standard input}: Assembler messages:
{standard input}:53: Error: bad instruction `svc #0'
2. remove unsupport compile option
cc1: error: unrecognized command line option "-mlittle-endian"
Signed-off-by: chao an <anchao@xiaomi.com>
Reference:
https://developer.arm.com/documentation/dui0474/m/image-structure-and-generation/section-placement-with-the-linker/section-placement-with-the-first-and-last-attributes
CAUTION:
FIRST and LAST must not violate the basic attribute sorting order. For example, FIRST RW is placed after any read-only code or read-only data.
arm-none-eabi-readelf -aS arm_vectors.o
1. Without const:
Section Headers:
[Nr] Name Type Addr Off Size ES Flg Lk Inf Al
[ 3] .vectors PROGBITS 00000000 000034 00011c 00 WA 0 0 4
2. const symbol:
[ 3] .vectors PROGBITS 00000000 000034 00011c 00 A 0 0 4
Regression by:
| commit 229b57d6cb
|
| arch/armv[6|7|8]-m: Move _vectors to arm_internal.h to avoid the duplication
|
| and change the type of _vectors from uint32_t to const void *
|
| Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Signed-off-by: chao an <anchao@xiaomi.com>
In file included from ./armv7-r/arm_l2cc_pl310.c:41:
./armv7-r/l2cc_pl310.h:38:10: fatal error: chip/chip.h: No such file or directory
38 | #include "chip/chip.h"
| ^~~~~~~~~~~~~
Signed-off-by: chao an <anchao@xiaomi.com>
stderr is a file stream pointer not a function name. Having
it in the renaming list may introduce linking error.
Signed-off-by: gaojiawei <gaojiawei@xiaomi.com>
* tools/configure.sh:
(USAGE): Explain both styles of usage:
- <board-name>:<config-name> for in-tree boards
- path to custom out-of-tree boards (relative to TOPDIR
or absolute)
Suggested by TimH in mailing list discussion "build board from custom
directory" started 15 Sept 2022, archived:
https://lists.apache.org/thread/7t8k79mm4kxy9cbo7vmybd36nzh94qtd
I noticed that the conn instance will leak during stress test,
The close work queued from tcp_close_eventhandler() will be canceled
by tcp_timer() immediately:
Breakpoint 1, tcp_close_eventhandler (dev=0x565cd338 <up_irq_restore+108>, pvpriv=0x5655e6ff <getpid+12>, flags=0) at tcp/tcp_close.c:71
(gdb) bt
| #0 tcp_close_eventhandler (dev=0x565cd338 <up_irq_restore+108>, pvpriv=0x5655e6ff <getpid+12>, flags=0) at tcp/tcp_close.c:71
| #1 0x5658bf1e in devif_conn_event (dev=0x5660bd80 <g_sim_dev>, flags=512, list=0x5660d558 <g_cbprealloc+312>) at devif/devif_callback.c:508
| #2 0x5658a219 in tcp_callback (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>, flags=512) at tcp/tcp_callback.c:167
| #3 0x56589253 in tcp_timer (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:378
| #4 0x5658dd47 in tcp_poll (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_devpoll.c:95
| #5 0x5658b95f in devif_poll_tcp_connections (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:601
| #6 0x5658b9ea in devif_poll (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:722
| #7 0x56577230 in netdriver_txavail_work (arg=0x5660bd80 <g_sim_dev>) at sim/up_netdriver.c:308
| #8 0x5655999e in work_thread (argc=2, argv=0xf3db5dd0) at wqueue/kwork_thread.c:178
| #9 0x5655983f in nxtask_start () at task/task_start.c:129
(gdb) c
Continuing.
Breakpoint 2, tcp_update_timer (conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:178
(gdb) bt
| #0 tcp_update_timer (conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:178
| #1 0x5658952a in tcp_timer (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:708
| #2 0x5658dd47 in tcp_poll (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_devpoll.c:95
| #3 0x5658b95f in devif_poll_tcp_connections (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:601
| #4 0x5658b9ea in devif_poll (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:722
| #5 0x56577230 in netdriver_txavail_work (arg=0x5660bd80 <g_sim_dev>) at sim/up_netdriver.c:308
| #6 0x5655999e in work_thread (argc=2, argv=0xf3db5dd0) at wqueue/kwork_thread.c:178
| #7 0x5655983f in nxtask_start () at task/task_start.c:129
Since a separate work will add 24 bytes to each conn instance,
but in order to support the feature of asynchronous close(),
I can not find a better way than adding a separate work,
for resource constraints, I recommend the developers to enable
CONFIG_NET_ALLOC_CONNS, which will reduce the ram usage.
Signed-off-by: chao an <anchao@xiaomi.com>
