Re:Re: [PATCH v4] powerpc:85xx: Add missing of_node_put() in sgy_cst1000
From: Liang He
Date: Fri Jun 17 2022 - 07:50:07 EST
在 2022-06-17 18:57:36，"Christophe Leroy" <christophe.leroy@xxxxxxxxxx> 写道：
>
>
>Le 17/06/2022 à 12:47, Liang He a écrit :
>>
>>
>>
>> At 2022-06-17 16:27:03, Conor.Dooley@xxxxxxxxxxxxx wrote:
>>> On 17/06/2022 09:17, Liang He wrote:
>>>>
>>>>
>>>>
>>>> At 2022-06-17 14:53:13, "Christophe Leroy" <christophe.leroy@xxxxxxxxxx> wrote:
>>>>>
>>>>>
>>>>> Le 17/06/2022 à 08:45, Liang He a écrit :
>>>>>>
>>>>>>
>>>>>>
>>>>>> At 2022-06-17 14:28:56, "Christophe Leroy" <christophe.leroy@xxxxxxxxxx> wrote:
>>>>>>>
>>>>>>>
>>>>>>> Le 17/06/2022 à 08:08, Liang He a écrit :
>>>>>>>> In gpio_halt_probe(), of_find_matching_node() will return a node
>>>>>>>> pointer with refcount incremented. We should use of_node_put() in
>>>>>>>> fail path or when it is not used anymore.
>>>>>>>>
>>>>>>>> Signed-off-by: Liang He <windhl@xxxxxxx>
>>>>>>>> ---
>>>>>>>> changelog:
>>>>>>>> v4: reuse exist 'err' and use a simple code style, advised by CJ
>>>>>>>> v3: use local 'child_node' advised by Michael.
>>>>>>>> v2: use goto-label patch style advised by Christophe Leroy.
>>>>>>>> v1: add of_node_put() before each exit.
>>>>>>>>
>>>>>>>> arch/powerpc/platforms/85xx/sgy_cts1000.c | 35 ++++++++++++++---------
>>>>>>>> 1 file changed, 22 insertions(+), 13 deletions(-)
>>>>>>>>
>>>>>>>> diff --git a/arch/powerpc/platforms/85xx/sgy_cts1000.c b/arch/powerpc/platforms/85xx/sgy_cts1000.c
>>>>>>>> index 98ae64075193..e4588943fe7e 100644
>>>>>>>> --- a/arch/powerpc/platforms/85xx/sgy_cts1000.c
>>>>>>>> +++ b/arch/powerpc/platforms/85xx/sgy_cts1000.c
>>>>>>>> @@ -71,6 +71,7 @@ static int gpio_halt_probe(struct platform_device *pdev)
>>>>>>>> {
>>>>>>>> enum of_gpio_flags flags;
>>>>>>>> struct device_node *node = pdev->dev.of_node;
>>>>>>>> + struct device_node *child_node;
>>>>>>>> int gpio, err, irq;
>>>>>>>> int trigger;
>>>>>>>>
>>>>>>>> @@ -78,26 +79,29 @@ static int gpio_halt_probe(struct platform_device *pdev)
>>>>>>>> return -ENODEV;
>>>>>>>>
>>>>>>>> /* If there's no matching child, this isn't really an error */
>>>>>>>> - halt_node = of_find_matching_node(node, child_match);
>>>>>>>> - if (!halt_node)
>>>>>>>> + child_node = of_find_matching_node(node, child_match);
>>>>>>>> + if (!child_node)
>>>>>>>> return 0;
>>>>>>>>
>>>>>>>> /* Technically we could just read the first one, but punish
>>>>>>>> * DT writers for invalid form. */
>>>>>>>> - if (of_gpio_count(halt_node) != 1)
>>>>>>>> - return -EINVAL;
>>>>>>>> + if (of_gpio_count(child_node) != 1) {
>>>>>>>> + err = -EINVAL;
>>>>>>>> + goto err_put;
>>>>>>>> + }
>>>>>>>>
>>>>>>>> /* Get the gpio number relative to the dynamic base. */
>>>>>>>> - gpio = of_get_gpio_flags(halt_node, 0, &flags);
>>>>>>>> - if (!gpio_is_valid(gpio))
>>>>>>>> - return -EINVAL;
>>>>>>>> + gpio = of_get_gpio_flags(child_node, 0, &flags);
>>>>>>>> + if (!gpio_is_valid(gpio)) {
>>>>>>>> + err = -EINVAL;
>>>>>>>> + gotot err_put;
>>>>>>>
>>>>>>> Did you test the build ?
>>>>>>
>>>>>> Sorry for this fault.
>>>>>>
>>>>>> In fact, I am still finding an efficient way to building different arch source code as I only have x86-64.
>>>>>>
>>>>>> Now I am try using QEMU.
>>>>>>
>>>>>> Anyway, sorry for this fault.
>>>>>
>>>>> You can find cross compilers for most architectures for x86-64 here :
>>>>> https://mirrors.edge.kernel.org/pub/tools/crosstool/
>>>>>
>>>>> Christophe
>>>>
>>>> Hi, Christophe and Conor.
>>>>
>>>> Sorry to trouble you again.
>>>>
>>>> Now I only know how to quickly identify the refcounting bugs, but I cannot efficiently give a build test.
>>>>
>>>> For example, I use the cross compilers 'powerpc-linux-gnu-gcc' to compile 'arch/powerpc/platforms/85xx/sgy_cts1000.c' with -fsyntax-only flag.
>>>> But I meet too many header file missing errors. Even if I add some 'include' pathes, e.g., ./arch/powerpc/include, ./include,
>>>> there are still too many other errors.
>>>>
>>>> So if there is any efficient way to check my patch code to avoid 'gotot' error again.
>>>
>>> idk anything about powerpc, but what I find is a nice way to get a compiler
>>> for an arch I don't use is to search on lore.kernel.org for a 0day robot
>>> build error since it gives instructions for building on that arch.
>>> For example:
>>> https://lore.kernel.org/linuxppc-dev/202206060910.rYNTFqdI-lkp@xxxxxxxxx/
>>>
>>>
>>> In this case, your bug seems obvious? You typed "gotot" instead of "goto".
>>>
>>> Hope that helps,
>>> Conor.
>>>
>>>>
>>>> Thanks again, Christophe and Conor.
>>>>
>>>> Liang
>>
>> Thanks, Conor and Christophe.
>>
>> I finally figure out an efficient way in which I can use cross-compiler to check my single patched file as follow:
>>
>> powerpc64le-linux-gnu-gcc -Wp,-MMD,arch/powerpc/kernel/.io.o.d -nostdinc -I./arch/powerpc/include -I./arch/powerpc/include/generated -I./include -I./arch/powerpc/include/uapi -I./arch/powerpc/include/generated/uapi -I./include/uapi -I./include/generated/uapi -include ./include/linux/compiler-version.h -include ./include/linux/kconfig.h -include ./include/linux/compiler_types.h -D__KERNEL__ -I ./arch/powerpc -fmacro-prefix-map=./= -Wall -Wundef -Werror=strict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common -fshort-wchar -fno-PIE -Werror=implicit-function-declaration -Werror=implicit-int -Werror=return-type -Wno-format-security -std=gnu11 -mcpu=powerpc -mcpu=powerpc -m32 -msoft-float -pipe -ffixed-r2 -mno-readonly-in-sdata -mno-altivec -mno-vsx -fno-asynchronous-unwind-tables -mno-string -mbig-endian -mstack-protector-guard=tls -mstack-protector-guard-reg=r2 -fno-delete-null-pointer-checks -Wno-frame-address -Wno-format-truncation -Wno-format-overflow -Wno-address-of-packed-member -O2 --param=allow-store-data-races=0 -Wframe-larger-than=1024 -fstack-protector-strong -Wimplicit-fallthrough=5 -Wno-main -Wno-unused-but-set-variable -Wno-unused-const-variable -fomit-frame-pointer -fno-stack-clash-protection -Wdeclaration-after-statement -Wvla -Wno-pointer-sign -Wcast-function-type -Wno-stringop-truncation -Wno-stringop-overflow -Wno-restrict -Wno-maybe-uninitialized -Wno-alloc-size-larger-than -fno-strict-overflow -fno-stack-check -fconserve-stack -Werror=date-time -Werror=incompatible-pointer-types -Werror=designated-init -Wno-packed-not-aligned -mstack-protector-guard-offset=576 -Werror -DKBUILD_MODNAME='"85xx"' -DKBUILD_MODFILE='"arch/powerpc/platforms/85xx/sgy_cts1000.c"' -fsyntax-only ./arch/powerpc/platforms/85xx/sgy_cts1000.c
>>
>
>Easiest way is:
>- Download cross compiler and extract it somewhere on your PC and ensure
>it is in the PATH.
>- Then:
>
>run "make CROSS_COMPILE=powerpc64-linux-gnu ARCH=powerpc
>corenet32_smp_defconfig"
>
>modify .config to activate CONFIG_SGY_CTS1000
>
>run "make CROSS_COMPILE=powerpc64-linux-gnu ARCH=powerpc
>arch/powerpc/platforms/85xx/sgy_cts1000.o"
>
Thanks for your valuable experience, Christophe.
It is great and I have used it to successfully compile my patched code without
any error:
"
CC arch/powerpc/platforms/85xx/sgy_cts1000.o
"
BYW, it should be 'make CROSS_COMPILE=powerpc64-linux-gnu- ...'
Thanks again!
Liang