* [PATCH] powerpc/lib: fix book3s/32 boot failure due to code patching
From: Christophe Leroy @ 2019-05-15 6:39 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, Greg KH,
erhard_f, Michael Neuling
Cc: linuxppc-dev, linux-kernel, stable
[Backport of upstream commit b45ba4a51cde29b2939365ef0c07ad34c8321789]
On powerpc32, patch_instruction() is called by apply_feature_fixups()
which is called from early_init()
There is the following note in front of early_init():
* Note that the kernel may be running at an address which is different
* from the address that it was linked at, so we must use RELOC/PTRRELOC
* to access static data (including strings). -- paulus
Therefore init_mem_is_free must be accessed with PTRRELOC()
Fixes: 1c38a84d4586 ("powerpc: Avoid code patching freed init sections")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=203597
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
---
Can't apply the upstream commit as such due to several other unrelated stuff
like for instance STRICT_KERNEL_RWX which are missing.
So instead, using same approach as for commit 252eb55816a6f69ef9464cad303cdb3326cdc61d
---
arch/powerpc/lib/code-patching.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/powerpc/lib/code-patching.c b/arch/powerpc/lib/code-patching.c
index 14535ad4cdd1..c312955977ce 100644
--- a/arch/powerpc/lib/code-patching.c
+++ b/arch/powerpc/lib/code-patching.c
@@ -23,7 +23,7 @@ int patch_instruction(unsigned int *addr, unsigned int instr)
int err;
/* Make sure we aren't patching a freed init section */
- if (init_mem_is_free && init_section_contains(addr, 4)) {
+ if (*PTRRELOC(&init_mem_is_free) && init_section_contains(addr, 4)) {
pr_debug("Skipping init section patching addr: 0x%px\n", addr);
return 0;
}
--
2.13.3
^ permalink raw reply related
* [PATCH stable 4.9] powerpc/lib: fix book3s/32 boot failure due to code patching
From: Christophe Leroy @ 2019-05-15 6:40 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, Greg KH,
erhard_f, Michael Neuling
Cc: linuxppc-dev, linux-kernel, stable
[Backport of upstream commit b45ba4a51cde29b2939365ef0c07ad34c8321789]
On powerpc32, patch_instruction() is called by apply_feature_fixups()
which is called from early_init()
There is the following note in front of early_init():
* Note that the kernel may be running at an address which is different
* from the address that it was linked at, so we must use RELOC/PTRRELOC
* to access static data (including strings). -- paulus
Therefore init_mem_is_free must be accessed with PTRRELOC()
Fixes: 1c38a84d4586 ("powerpc: Avoid code patching freed init sections")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=203597
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
---
Can't apply the upstream commit as such due to several other unrelated stuff
like for instance STRICT_KERNEL_RWX which are missing.
So instead, using same approach as for commit 252eb55816a6f69ef9464cad303cdb3326cdc61d
---
arch/powerpc/lib/code-patching.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/powerpc/lib/code-patching.c b/arch/powerpc/lib/code-patching.c
index 14535ad4cdd1..c312955977ce 100644
--- a/arch/powerpc/lib/code-patching.c
+++ b/arch/powerpc/lib/code-patching.c
@@ -23,7 +23,7 @@ int patch_instruction(unsigned int *addr, unsigned int instr)
int err;
/* Make sure we aren't patching a freed init section */
- if (init_mem_is_free && init_section_contains(addr, 4)) {
+ if (*PTRRELOC(&init_mem_is_free) && init_section_contains(addr, 4)) {
pr_debug("Skipping init section patching addr: 0x%px\n", addr);
return 0;
}
--
2.13.3
^ permalink raw reply related
* Re: [PATCH] powerpc/mm/book3s64: Implement STRICT_MODULE_RWX
From: Christoph Hellwig @ 2019-05-15 6:41 UTC (permalink / raw)
To: Russell Currey; +Cc: linuxppc-dev
In-Reply-To: <20190515013000.16085-1-ruscur@russell.cc>
> + * This program is free software; you can redistribute it and/or modify it
> + * under the terms of the GNU General Public License as published by the
> + * Free Software Foundation; either version 2 of the License, or (at your
> + * option) any later version.
This license boilerplate should not be added together with an SPDX tag.
> +// we need this to have a single pointer to pass into apply_to_page_range()
Please use normal /* - */ style comments.
^ permalink raw reply
* Re: [PATCH] powerpc/lib: fix book3s/32 boot failure due to code patching
From: Christophe Leroy @ 2019-05-15 6:41 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, Greg KH,
erhard_f, Michael Neuling
Cc: linuxppc-dev, linux-kernel, stable
In-Reply-To: <629c2acb1fcd09c2d2e3352370c3d9853372cf39.1557902321.git.christophe.leroy@c-s.fr>
Oops, forgot to tell it's for 4.9. Resending with proper subject.
Le 15/05/2019 à 08:39, Christophe Leroy a écrit :
> [Backport of upstream commit b45ba4a51cde29b2939365ef0c07ad34c8321789]
>
> On powerpc32, patch_instruction() is called by apply_feature_fixups()
> which is called from early_init()
>
> There is the following note in front of early_init():
> * Note that the kernel may be running at an address which is different
> * from the address that it was linked at, so we must use RELOC/PTRRELOC
> * to access static data (including strings). -- paulus
>
> Therefore init_mem_is_free must be accessed with PTRRELOC()
>
> Fixes: 1c38a84d4586 ("powerpc: Avoid code patching freed init sections")
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=203597
> Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
>
> ---
> Can't apply the upstream commit as such due to several other unrelated stuff
> like for instance STRICT_KERNEL_RWX which are missing.
> So instead, using same approach as for commit 252eb55816a6f69ef9464cad303cdb3326cdc61d
> ---
> arch/powerpc/lib/code-patching.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/powerpc/lib/code-patching.c b/arch/powerpc/lib/code-patching.c
> index 14535ad4cdd1..c312955977ce 100644
> --- a/arch/powerpc/lib/code-patching.c
> +++ b/arch/powerpc/lib/code-patching.c
> @@ -23,7 +23,7 @@ int patch_instruction(unsigned int *addr, unsigned int instr)
> int err;
>
> /* Make sure we aren't patching a freed init section */
> - if (init_mem_is_free && init_section_contains(addr, 4)) {
> + if (*PTRRELOC(&init_mem_is_free) && init_section_contains(addr, 4)) {
> pr_debug("Skipping init section patching addr: 0x%px\n", addr);
> return 0;
> }
>
^ permalink raw reply
* [PATCH RESEND V6 0/3] Support more sample rate in asrc
From: S.j. Wang @ 2019-05-15 6:42 UTC (permalink / raw)
To: timur@kernel.org, nicoleotsuka@gmail.com, Xiubo.Lee@gmail.com,
festevam@gmail.com, broonie@kernel.org,
alsa-devel@alsa-project.org
Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org
Support more sample rate in asrc
Shengjiu Wang (3):
ASoC: fsl_asrc: Fix the issue about unsupported rate
ASoC: fsl_asrc: replace the process_option table with function
ASoC: fsl_asrc: Unify the supported input and output rate
Changes in RESEND V6
- change the Content-Transfer-Encoding to "quoted-printable", for
- "base64" can't be applied
Changes in v6
- add acked-by
- fixed minor issue according to comments in v5
Changes in v5
- fix the [1/24, 8]
- move fsl_asrc_sel_proc before setting
Changes in v4
- add patch to Fix the [8kHz, 30kHz] open set issue.
Changes in v3
- remove FSL_ASRC_RATES
- refine fsl_asrc_sel_proc according to comments
Changes in v2
- add more comments in code
- add commit "Unify the supported input and output rate"
sound/soc/fsl/fsl_asrc.c | 105 ++++++++++++++++++++++++++-------------
1 file changed, 71 insertions(+), 34 deletions(-)
--
2.21.0
^ permalink raw reply
* [PATCH RESEND V6 1/3] ASoC: fsl_asrc: Fix the issue about unsupported rate
From: S.j. Wang @ 2019-05-15 6:42 UTC (permalink / raw)
To: timur@kernel.org, nicoleotsuka@gmail.com, Xiubo.Lee@gmail.com,
festevam@gmail.com, broonie@kernel.org,
alsa-devel@alsa-project.org
Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org
In-Reply-To: <cover.1557901312.git.shengjiu.wang@nxp.com>
When the output sample rate is [8kHz, 30kHz], the limitation
of the supported ratio range is [1/24, 8]. In the driver
we use (8kHz, 30kHz) instead of [8kHz, 30kHz].
So this patch is to fix this issue and the potential rounding
issue with divider.
Fixes: fff6e03c7b65 ("ASoC: fsl_asrc: add support for 8-30kHz
output sample rate")
Cc: <stable@vger.kernel.org>
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
---
sound/soc/fsl/fsl_asrc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
index 0b937924d2e4..ea035c12a325 100644
--- a/sound/soc/fsl/fsl_asrc.c
+++ b/sound/soc/fsl/fsl_asrc.c
@@ -282,8 +282,8 @@ static int fsl_asrc_config_pair(struct fsl_asrc_pair *pair)
return -EINVAL;
}
- if ((outrate > 8000 && outrate < 30000) &&
- (outrate/inrate > 24 || inrate/outrate > 8)) {
+ if ((outrate >= 8000 && outrate <= 30000) &&
+ (outrate > 24 * inrate || inrate > 8 * outrate)) {
pair_err("exceed supported ratio range [1/24, 8] for \
inrate/outrate: %d/%d\n", inrate, outrate);
return -EINVAL;
--
2.21.0
^ permalink raw reply related
* [PATCH RESEND V6 2/3] ASoC: fsl_asrc: replace the process_option table with function
From: S.j. Wang @ 2019-05-15 6:42 UTC (permalink / raw)
To: timur@kernel.org, nicoleotsuka@gmail.com, Xiubo.Lee@gmail.com,
festevam@gmail.com, broonie@kernel.org,
alsa-devel@alsa-project.org
Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org
In-Reply-To: <cover.1557901312.git.shengjiu.wang@nxp.com>
When we want to support more sample rate, for example 12kHz/24kHz
we need update the process_option table, if we want to support more
sample rate next time, the table need to be updated again. which
is not flexible.
We got a function fsl_asrc_sel_proc to replace the table, which can
give the pre-processing and post-processing options according to
the sample rate.
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
---
sound/soc/fsl/fsl_asrc.c | 71 +++++++++++++++++++++++++++++-----------
1 file changed, 51 insertions(+), 20 deletions(-)
diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
index ea035c12a325..a8d6710f2541 100644
--- a/sound/soc/fsl/fsl_asrc.c
+++ b/sound/soc/fsl/fsl_asrc.c
@@ -26,24 +26,6 @@
#define pair_dbg(fmt, ...) \
dev_dbg(&asrc_priv->pdev->dev, "Pair %c: " fmt, 'A' + index, ##__VA_ARGS__)
-/* Sample rates are aligned with that defined in pcm.h file */
-static const u8 process_option[][12][2] = {
- /* 8kHz 11.025kHz 16kHz 22.05kHz 32kHz 44.1kHz 48kHz 64kHz 88.2kHz 96kHz 176kHz 192kHz */
- {{0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0},}, /* 5512Hz */
- {{0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0},}, /* 8kHz */
- {{0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0},}, /* 11025Hz */
- {{1, 2}, {0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0},}, /* 16kHz */
- {{1, 2}, {1, 2}, {0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0}, {0, 0}, {0, 0}, {0, 0},}, /* 22050Hz */
- {{1, 2}, {2, 1}, {2, 1}, {0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0}, {0, 0},}, /* 32kHz */
- {{2, 2}, {2, 2}, {2, 1}, {2, 1}, {0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0},}, /* 44.1kHz */
- {{2, 2}, {2, 2}, {2, 1}, {2, 1}, {0, 2}, {0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0}, {0, 0},}, /* 48kHz */
- {{2, 2}, {2, 2}, {2, 2}, {2, 1}, {1, 2}, {0, 2}, {0, 2}, {0, 1}, {0, 1}, {0, 1}, {0, 1}, {0, 0},}, /* 64kHz */
- {{2, 2}, {2, 2}, {2, 2}, {2, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 1}, {1, 1}, {1, 1}, {1, 1}, {1, 1},}, /* 88.2kHz */
- {{2, 2}, {2, 2}, {2, 2}, {2, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 1}, {1, 1}, {1, 1}, {1, 1}, {1, 1},}, /* 96kHz */
- {{2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 1}, {2, 1}, {2, 1}, {2, 1}, {2, 1},}, /* 176kHz */
- {{2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 1}, {2, 1}, {2, 1}, {2, 1}, {2, 1},}, /* 192kHz */
-};
-
/* Corresponding to process_option */
static int supported_input_rate[] = {
5512, 8000, 11025, 16000, 22050, 32000, 44100, 48000, 64000, 88200,
@@ -79,6 +61,52 @@ static unsigned char output_clk_map_imx53[] = {
static unsigned char *clk_map[2];
+/**
+ * Select the pre-processing and post-processing options
+ * Make sure to exclude following unsupported cases before
+ * calling this function:
+ * 1) inrate > 8.125 * outrate
+ * 2) inrate > 16.125 * outrate
+ *
+ * inrate: input sample rate
+ * outrate: output sample rate
+ * pre_proc: return value for pre-processing option
+ * post_proc: return value for post-processing option
+ */
+static void fsl_asrc_sel_proc(int inrate, int outrate,
+ int *pre_proc, int *post_proc)
+{
+ bool post_proc_cond2;
+ bool post_proc_cond0;
+
+ /* select pre_proc between [0, 2] */
+ if (inrate * 8 > 33 * outrate)
+ *pre_proc = 2;
+ else if (inrate * 8 > 15 * outrate) {
+ if (inrate > 152000)
+ *pre_proc = 2;
+ else
+ *pre_proc = 1;
+ } else if (inrate < 76000)
+ *pre_proc = 0;
+ else if (inrate > 152000)
+ *pre_proc = 2;
+ else
+ *pre_proc = 1;
+
+ /* Condition for selection of post-processing */
+ post_proc_cond2 = (inrate * 15 > outrate * 16 && outrate < 56000) ||
+ (inrate > 56000 && outrate < 56000);
+ post_proc_cond0 = inrate * 23 < outrate * 8;
+
+ if (post_proc_cond2)
+ *post_proc = 2;
+ else if (post_proc_cond0)
+ *post_proc = 0;
+ else
+ *post_proc = 1;
+}
+
/**
* Request ASRC pair
*
@@ -239,6 +267,7 @@ static int fsl_asrc_config_pair(struct fsl_asrc_pair *pair)
u32 inrate, outrate, indiv, outdiv;
u32 clk_index[2], div[2];
int in, out, channels;
+ int pre_proc, post_proc;
struct clk *clk;
bool ideal;
@@ -377,11 +406,13 @@ static int fsl_asrc_config_pair(struct fsl_asrc_pair *pair)
ASRCTR_IDRi_MASK(index) | ASRCTR_USRi_MASK(index),
ASRCTR_IDR(index) | ASRCTR_USR(index));
+ fsl_asrc_sel_proc(inrate, outrate, &pre_proc, &post_proc);
+
/* Apply configurations for pre- and post-processing */
regmap_update_bits(asrc_priv->regmap, REG_ASRCFG,
ASRCFG_PREMODi_MASK(index) | ASRCFG_POSTMODi_MASK(index),
- ASRCFG_PREMOD(index, process_option[in][out][0]) |
- ASRCFG_POSTMOD(index, process_option[in][out][1]));
+ ASRCFG_PREMOD(index, pre_proc) |
+ ASRCFG_POSTMOD(index, post_proc));
return fsl_asrc_set_ideal_ratio(pair, inrate, outrate);
}
--
2.21.0
^ permalink raw reply related
* [PATCH RESEND V6 3/3] ASoC: fsl_asrc: Unify the supported input and output rate
From: S.j. Wang @ 2019-05-15 6:42 UTC (permalink / raw)
To: timur@kernel.org, nicoleotsuka@gmail.com, Xiubo.Lee@gmail.com,
festevam@gmail.com, broonie@kernel.org,
alsa-devel@alsa-project.org
Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org
In-Reply-To: <cover.1557901312.git.shengjiu.wang@nxp.com>
Unify the supported input and output rate, add the
12kHz/24kHz/128kHz to the support list
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
---
sound/soc/fsl/fsl_asrc.c | 32 +++++++++++++++++++-------------
1 file changed, 19 insertions(+), 13 deletions(-)
diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
index a8d6710f2541..cbbf6257f08a 100644
--- a/sound/soc/fsl/fsl_asrc.c
+++ b/sound/soc/fsl/fsl_asrc.c
@@ -27,13 +27,14 @@
dev_dbg(&asrc_priv->pdev->dev, "Pair %c: " fmt, 'A' + index, ##__VA_ARGS__)
/* Corresponding to process_option */
-static int supported_input_rate[] = {
- 5512, 8000, 11025, 16000, 22050, 32000, 44100, 48000, 64000, 88200,
- 96000, 176400, 192000,
+static unsigned int supported_asrc_rate[] = {
+ 5512, 8000, 11025, 12000, 16000, 22050, 24000, 32000, 44100, 48000,
+ 64000, 88200, 96000, 128000, 176400, 192000,
};
-static int supported_asrc_rate[] = {
- 8000, 11025, 16000, 22050, 32000, 44100, 48000, 64000, 88200, 96000, 176400, 192000,
+static struct snd_pcm_hw_constraint_list fsl_asrc_rate_constraints = {
+ .count = ARRAY_SIZE(supported_asrc_rate),
+ .list = supported_asrc_rate,
};
/**
@@ -293,11 +294,11 @@ static int fsl_asrc_config_pair(struct fsl_asrc_pair *pair)
ideal = config->inclk == INCLK_NONE;
/* Validate input and output sample rates */
- for (in = 0; in < ARRAY_SIZE(supported_input_rate); in++)
- if (inrate == supported_input_rate[in])
+ for (in = 0; in < ARRAY_SIZE(supported_asrc_rate); in++)
+ if (inrate == supported_asrc_rate[in])
break;
- if (in == ARRAY_SIZE(supported_input_rate)) {
+ if (in == ARRAY_SIZE(supported_asrc_rate)) {
pair_err("unsupported input sample rate: %dHz\n", inrate);
return -EINVAL;
}
@@ -311,7 +312,7 @@ static int fsl_asrc_config_pair(struct fsl_asrc_pair *pair)
return -EINVAL;
}
- if ((outrate >= 8000 && outrate <= 30000) &&
+ if ((outrate >= 5512 && outrate <= 30000) &&
(outrate > 24 * inrate || inrate > 8 * outrate)) {
pair_err("exceed supported ratio range [1/24, 8] for \
inrate/outrate: %d/%d\n", inrate, outrate);
@@ -486,7 +487,9 @@ static int fsl_asrc_dai_startup(struct snd_pcm_substream *substream,
snd_pcm_hw_constraint_step(substream->runtime, 0,
SNDRV_PCM_HW_PARAM_CHANNELS, 2);
- return 0;
+
+ return snd_pcm_hw_constraint_list(substream->runtime, 0,
+ SNDRV_PCM_HW_PARAM_RATE, &fsl_asrc_rate_constraints);
}
static int fsl_asrc_dai_hw_params(struct snd_pcm_substream *substream,
@@ -599,7 +602,6 @@ static int fsl_asrc_dai_probe(struct snd_soc_dai *dai)
return 0;
}
-#define FSL_ASRC_RATES SNDRV_PCM_RATE_8000_192000
#define FSL_ASRC_FORMATS (SNDRV_PCM_FMTBIT_S24_LE | \
SNDRV_PCM_FMTBIT_S16_LE | \
SNDRV_PCM_FMTBIT_S20_3LE)
@@ -610,14 +612,18 @@ static struct snd_soc_dai_driver fsl_asrc_dai = {
.stream_name = "ASRC-Playback",
.channels_min = 1,
.channels_max = 10,
- .rates = FSL_ASRC_RATES,
+ .rate_min = 5512,
+ .rate_max = 192000,
+ .rates = SNDRV_PCM_RATE_KNOT,
.formats = FSL_ASRC_FORMATS,
},
.capture = {
.stream_name = "ASRC-Capture",
.channels_min = 1,
.channels_max = 10,
- .rates = FSL_ASRC_RATES,
+ .rate_min = 5512,
+ .rate_max = 192000,
+ .rates = SNDRV_PCM_RATE_KNOT,
.formats = FSL_ASRC_FORMATS,
},
.ops = &fsl_asrc_dai_ops,
--
2.21.0
^ permalink raw reply related
* Re: [RFC PATCH] powerpc/mm: Implement STRICT_MODULE_RWX
From: Christoph Hellwig @ 2019-05-15 6:42 UTC (permalink / raw)
To: Christophe Leroy; +Cc: linux-kernel, Paul Mackerras, linuxppc-dev
In-Reply-To: <df502ffe07caa38c46b0144fc824fff447f4105b.1557901092.git.christophe.leroy@c-s.fr>
> +static int change_page_ro(pte_t *ptep, pgtable_t token, unsigned long addr, void *data)
There are a couple way too long lines like this in the patch.
^ permalink raw reply
* Re: [RFC PATCH] powerpc/mm: Implement STRICT_MODULE_RWX
From: Christophe Leroy @ 2019-05-15 6:44 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: linux-kernel, Paul Mackerras, linuxppc-dev
In-Reply-To: <20190515064205.GB15778@infradead.org>
Le 15/05/2019 à 08:42, Christoph Hellwig a écrit :
>> +static int change_page_ro(pte_t *ptep, pgtable_t token, unsigned long addr, void *data)
>
> There are a couple way too long lines like this in the patch.
>
powerpc arch accepts 90 chars per line, see arch/powerpc/tools/checkpatch.pl
Christophe
^ permalink raw reply
* Re: [PATCH, RFC] byteorder: sanity check toolchain vs kernel endianess
From: Arnd Bergmann @ 2019-05-15 6:53 UTC (permalink / raw)
To: Christoph Hellwig
Cc: linux-arch, linuxppc-dev, Linux Kernel Mailing List,
Andrew Donnellan, Nick Kossifidis, Andrew Morton, Linus Torvalds,
Dmitry Vyukov
In-Reply-To: <20190513120435.GB22993@lst.de>
On Mon, May 13, 2019 at 2:04 PM Christoph Hellwig <hch@lst.de> wrote:
>
> On Mon, May 13, 2019 at 01:50:19PM +0200, Dmitry Vyukov wrote:
> > > We did have some bugs in the past (~1-2 y/ago) but AFAIK they are all
> > > fixed now. These days I build most of my kernels with a bi-endian 64-bit
> > > toolchain, and switching endian without running `make clean` also works.
> >
> > For the record, yes, it turn out to be a problem in our code (a latent
> > bug). We actually used host (x86) gcc to build as-if ppc code that can
> > run on the host, so it defined neither LE no BE macros. It just
> > happened to work in the past :)
>
> So Nick was right and these checks actually are useful..
Yes, definitely. I wonder if we should also bring back the word size check
from include/asm-generic/bitsperlong.h, which was disabled right
after I originally added that.
Arnd
^ permalink raw reply
* Re: [PATCH] vsprintf: Do not break early boot with probing addresses
From: Geert Uytterhoeven @ 2019-05-15 7:23 UTC (permalink / raw)
To: Steven Rostedt
Cc: Petr Mladek, linux-arch@vger.kernel.org, Sergey Senozhatsky,
Heiko Carstens, linux-s390@vger.kernel.org,
linuxppc-dev@lists.ozlabs.org, Rasmus Villemoes,
linux-kernel@vger.kernel.org, Michal Hocko, Sergey Senozhatsky,
David Laight, Stephen Rothwell, Andy Shevchenko, Linus Torvalds,
Martin Schwidefsky, Tobin C . Harding
In-Reply-To: <20190514153503.6b7faaa7@oasis.local.home>
Hi Steve,
On Tue, May 14, 2019 at 9:35 PM Steven Rostedt <rostedt@goodmis.org> wrote:
> On Tue, 14 May 2019 21:13:06 +0200
> Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> > > > Do we care about the value? "(-E%u)"?
> > >
> > > That too could be confusing. What would (-E22) be considered by a user
> > > doing an sprintf() on some string. I know that would confuse me, or I
> > > would think that it was what the %pX displayed, and wonder why it
> > > displayed it that way. Whereas "(fault)" is quite obvious for any %p
> > > use case.
> >
> > I would immediately understand there's a missing IS_ERR() check in a
> > function that can return -EINVAL, without having to add a new printk()
> > to find out what kind of bogus value has been received, and without
> > having to reboot, and trying to reproduce...
>
> I have to ask. Has there actually been a case that you used a %pX and
> it faulted, and you had to go back to find what the value of the
> failure was?
If it faulted, the bad pointer value is obvious from the backtrace.
If the code avoids the fault by verifying the pointer and returning
"(efault)" instead, the bad pointer value is lost.
Or am I missing something?
Thanks!
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply
* Re: [PATCH] vsprintf: Do not break early boot with probing addresses
From: Petr Mladek @ 2019-05-15 7:35 UTC (permalink / raw)
To: Steven Rostedt
Cc: linux-arch@vger.kernel.org, Sergey Senozhatsky, Heiko Carstens,
linux-s390@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
Rasmus Villemoes, linux-kernel@vger.kernel.org, Michal Hocko,
Sergey Senozhatsky, David Laight, Geert Uytterhoeven,
Stephen Rothwell, Andy Shevchenko, Linus Torvalds,
Martin Schwidefsky, Tobin C . Harding
In-Reply-To: <20190514143751.48e81e05@oasis.local.home>
On Tue 2019-05-14 14:37:51, Steven Rostedt wrote:
>
> [ Purple is a nice shade on the bike shed. ;-) ]
>
> On Tue, 14 May 2019 11:02:17 +0200
> Geert Uytterhoeven <geert@linux-m68k.org> wrote:
>
> > On Tue, May 14, 2019 at 10:29 AM David Laight <David.Laight@aculab.com> wrote:
> > > > And I like Steven's "(fault)" idea.
> > > > How about this:
> > > >
> > > > if ptr < PAGE_SIZE -> "(null)"
> > > > if IS_ERR_VALUE(ptr) -> "(fault)"
> > > >
> > > > -ss
> > >
> > > Or:
> > > if (ptr < PAGE_SIZE)
> > > return ptr ? "(null+)" : "(null)";
>
> Hmm, that is useful.
>
> > > if IS_ERR_VALUE(ptr)
> > > return "(errno)"
>
> I still prefer "(fault)" as is pretty much all I would expect from a
> pointer dereference, even if it is just bad parsing of, say, a parsing
> an MAC address. "fault" is generic enough. "errno" will be confusing,
> because that's normally a variable not a output.
>
> >
> > Do we care about the value? "(-E%u)"?
>
> That too could be confusing. What would (-E22) be considered by a user
> doing an sprintf() on some string. I know that would confuse me, or I
> would think that it was what the %pX displayed, and wonder why it
> displayed it that way. Whereas "(fault)" is quite obvious for any %p
> use case.
This discussion clearly shows that it is hard to make anyone happy.
I considered switching to "(fault)" because there seems to be more
people in favor of this.
But there is used also "(einval)" when an unsupported pointer
modifier is passed. The idea is to show error codes that people
are familiar with.
It might have been better to use the uppercase "(EFAULT)" and
"(EINVAL)" to make it more obvious. But I wanted to follow
the existing style with the lowercase "(null)".
As of now, I think that we should keep it as is unless there is
some wider agreement on a change.
Best Regards,
Petr
^ permalink raw reply
* [PATCH 1/3] dt-bindings: pci: layerscape-pci: add compatible strings "fsl, ls1028a-pcie"
From: Xiaowei Bao @ 2019-05-15 7:27 UTC (permalink / raw)
To: bhelgaas, robh+dt, mark.rutland, shawnguo, leoyang.li, kishon,
lorenzo.pieralisi, arnd, gregkh, minghuan.Lian, mingkai.hu,
roy.zang, kstewart, pombredanne, shawn.lin, linux-pci, devicetree,
linux-kernel, linux-arm-kernel, linuxppc-dev
Cc: Xiaowei Bao
Add the PCIe compatible string for LS1028A
Signed-off-by: Xiaowei Bao <xiaowei.bao@nxp.com>
---
.../devicetree/bindings/pci/layerscape-pci.txt | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/Documentation/devicetree/bindings/pci/layerscape-pci.txt b/Documentation/devicetree/bindings/pci/layerscape-pci.txt
index e20ceaa..99a386e 100644
--- a/Documentation/devicetree/bindings/pci/layerscape-pci.txt
+++ b/Documentation/devicetree/bindings/pci/layerscape-pci.txt
@@ -21,6 +21,7 @@ Required properties:
"fsl,ls1046a-pcie"
"fsl,ls1043a-pcie"
"fsl,ls1012a-pcie"
+ "fsl,ls1028a-pcie"
EP mode:
"fsl,ls1046a-pcie-ep", "fsl,ls-pcie-ep"
- reg: base addresses and lengths of the PCIe controller register blocks.
--
1.7.1
^ permalink raw reply related
* [PATCH 2/3] arm64: dts: ls1028a: Add PCIe controller DT nodes
From: Xiaowei Bao @ 2019-05-15 7:27 UTC (permalink / raw)
To: bhelgaas, robh+dt, mark.rutland, shawnguo, leoyang.li, kishon,
lorenzo.pieralisi, arnd, gregkh, minghuan.Lian, mingkai.hu,
roy.zang, kstewart, pombredanne, shawn.lin, linux-pci, devicetree,
linux-kernel, linux-arm-kernel, linuxppc-dev
Cc: Xiaowei Bao
In-Reply-To: <20190515072747.39941-1-xiaowei.bao@nxp.com>
LS1028a implements 2 PCIe 3.0 controllers.
Signed-off-by: Xiaowei Bao <xiaowei.bao@nxp.com>
---
arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi | 52 ++++++++++++++++++++++++
1 files changed, 52 insertions(+), 0 deletions(-)
diff --git a/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi b/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
index b045812..50b579b 100644
--- a/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
+++ b/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
@@ -398,6 +398,58 @@
status = "disabled";
};
+ pcie@3400000 {
+ compatible = "fsl,ls1028a-pcie";
+ reg = <0x00 0x03400000 0x0 0x00100000 /* controller registers */
+ 0x80 0x00000000 0x0 0x00002000>; /* configuration space */
+ reg-names = "regs", "config";
+ interrupts = <GIC_SPI 108 IRQ_TYPE_LEVEL_HIGH>, /* PME interrupt */
+ <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>; /* aer interrupt */
+ interrupt-names = "pme", "aer";
+ #address-cells = <3>;
+ #size-cells = <2>;
+ device_type = "pci";
+ dma-coherent;
+ num-lanes = <4>;
+ bus-range = <0x0 0xff>;
+ ranges = <0x81000000 0x0 0x00000000 0x80 0x00010000 0x0 0x00010000 /* downstream I/O */
+ 0x82000000 0x0 0x40000000 0x80 0x40000000 0x0 0x40000000>; /* non-prefetchable memory */
+ msi-parent = <&its>;
+ #interrupt-cells = <1>;
+ interrupt-map-mask = <0 0 0 7>;
+ interrupt-map = <0000 0 0 1 &gic GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>,
+ <0000 0 0 2 &gic GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>,
+ <0000 0 0 3 &gic GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>,
+ <0000 0 0 4 &gic GIC_SPI 112 IRQ_TYPE_LEVEL_HIGH>;
+ status = "disabled";
+ };
+
+ pcie@3500000 {
+ compatible = "fsl,ls1028a-pcie";
+ reg = <0x00 0x03500000 0x0 0x00100000 /* controller registers */
+ 0x88 0x00000000 0x0 0x00002000>; /* configuration space */
+ reg-names = "regs", "config";
+ interrupts = <GIC_SPI 113 IRQ_TYPE_LEVEL_HIGH>,
+ <GIC_SPI 114 IRQ_TYPE_LEVEL_HIGH>;
+ interrupt-names = "pme", "aer";
+ #address-cells = <3>;
+ #size-cells = <2>;
+ device_type = "pci";
+ dma-coherent;
+ num-lanes = <4>;
+ bus-range = <0x0 0xff>;
+ ranges = <0x81000000 0x0 0x00000000 0x88 0x00010000 0x0 0x00010000 /* downstream I/O */
+ 0x82000000 0x0 0x40000000 0x88 0x40000000 0x0 0x40000000>; /* non-prefetchable memory */
+ msi-parent = <&its>;
+ #interrupt-cells = <1>;
+ interrupt-map-mask = <0 0 0 7>;
+ interrupt-map = <0000 0 0 1 &gic GIC_SPI 114 IRQ_TYPE_LEVEL_HIGH>,
+ <0000 0 0 2 &gic GIC_SPI 115 IRQ_TYPE_LEVEL_HIGH>,
+ <0000 0 0 3 &gic GIC_SPI 116 IRQ_TYPE_LEVEL_HIGH>,
+ <0000 0 0 4 &gic GIC_SPI 117 IRQ_TYPE_LEVEL_HIGH>;
+ status = "disabled";
+ };
+
pcie@1f0000000 { /* Integrated Endpoint Root Complex */
compatible = "pci-host-ecam-generic";
reg = <0x01 0xf0000000 0x0 0x100000>;
--
1.7.1
^ permalink raw reply related
* [PATCH 3/3] PCI: layerscape: Add LS1028a support
From: Xiaowei Bao @ 2019-05-15 7:27 UTC (permalink / raw)
To: bhelgaas, robh+dt, mark.rutland, shawnguo, leoyang.li, kishon,
lorenzo.pieralisi, arnd, gregkh, minghuan.Lian, mingkai.hu,
roy.zang, kstewart, pombredanne, shawn.lin, linux-pci, devicetree,
linux-kernel, linux-arm-kernel, linuxppc-dev
Cc: Xiaowei Bao
In-Reply-To: <20190515072747.39941-1-xiaowei.bao@nxp.com>
Add support for the LS1028a PCIe controller.
Signed-off-by: Xiaowei Bao <xiaowei.bao@nxp.com>
---
drivers/pci/controller/dwc/pci-layerscape.c | 9 +++++++++
1 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/drivers/pci/controller/dwc/pci-layerscape.c b/drivers/pci/controller/dwc/pci-layerscape.c
index 3a5fa26..8c556e1 100644
--- a/drivers/pci/controller/dwc/pci-layerscape.c
+++ b/drivers/pci/controller/dwc/pci-layerscape.c
@@ -236,6 +236,14 @@ static int ls_pcie_msi_host_init(struct pcie_port *pp)
.dw_pcie_ops = &dw_ls_pcie_ops,
};
+static const struct ls_pcie_drvdata ls1028a_drvdata = {
+ .lut_offset = 0x80000,
+ .ltssm_shift = 0,
+ .lut_dbg = 0x407fc,
+ .ops = &ls_pcie_host_ops,
+ .dw_pcie_ops = &dw_ls_pcie_ops,
+};
+
static const struct ls_pcie_drvdata ls1046_drvdata = {
.lut_offset = 0x80000,
.ltssm_shift = 24,
@@ -263,6 +271,7 @@ static int ls_pcie_msi_host_init(struct pcie_port *pp)
static const struct of_device_id ls_pcie_of_match[] = {
{ .compatible = "fsl,ls1012a-pcie", .data = &ls1046_drvdata },
{ .compatible = "fsl,ls1021a-pcie", .data = &ls1021_drvdata },
+ { .compatible = "fsl,ls1028a-pcie", .data = &ls1028a_drvdata },
{ .compatible = "fsl,ls1043a-pcie", .data = &ls1043_drvdata },
{ .compatible = "fsl,ls1046a-pcie", .data = &ls1046_drvdata },
{ .compatible = "fsl,ls2080a-pcie", .data = &ls2080_drvdata },
--
1.7.1
^ permalink raw reply related
* [PATCH v3] powerpc/pseries: Fix cpu_hotplug_lock acquisition in resize_hpt()
From: Gautham R. Shenoy @ 2019-05-15 7:45 UTC (permalink / raw)
To: Paul Mackerras, Nicholas Piggin, Michael Ellerman,
Aneesh Kumar K.V
Cc: Gautham R. Shenoy, linuxppc-dev, linux-kernel
From: "Gautham R. Shenoy" <ego@linux.vnet.ibm.com>
The calls to arch_add_memory()/arch_remove_memory() are always made
with the read-side cpu_hotplug_lock acquired via
memory_hotplug_begin(). On pSeries,
arch_add_memory()/arch_remove_memory() eventually call resize_hpt()
which in turn calls stop_machine() which acquires the read-side
cpu_hotplug_lock again, thereby resulting in the recursive acquisition
of this lock.
Lockdep complains as follows in these code-paths.
swapper/0/1 is trying to acquire lock:
(____ptrval____) (cpu_hotplug_lock.rw_sem){++++}, at: stop_machine+0x2c/0x60
but task is already holding lock:
(____ptrval____) (cpu_hotplug_lock.rw_sem){++++}, at: mem_hotplug_begin+0x20/0x50
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(cpu_hotplug_lock.rw_sem);
lock(cpu_hotplug_lock.rw_sem);
*** DEADLOCK ***
May be due to missing lock nesting notation
3 locks held by swapper/0/1:
#0: (____ptrval____) (&dev->mutex){....}, at: __driver_attach+0x12c/0x1b0
#1: (____ptrval____) (cpu_hotplug_lock.rw_sem){++++}, at: mem_hotplug_begin+0x20/0x50
#2: (____ptrval____) (mem_hotplug_lock.rw_sem){++++}, at: percpu_down_write+0x54/0x1a0
stack backtrace:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.0.0-rc5-58373-gbc99402235f3-dirty #166
Call Trace:
[c0000000feb03150] [c000000000e32bd4] dump_stack+0xe8/0x164 (unreliable)
[c0000000feb031a0] [c00000000020d6c0] __lock_acquire+0x1110/0x1c70
[c0000000feb03320] [c00000000020f080] lock_acquire+0x240/0x290
[c0000000feb033e0] [c00000000017f554] cpus_read_lock+0x64/0xf0
[c0000000feb03420] [c00000000029ebac] stop_machine+0x2c/0x60
[c0000000feb03460] [c0000000000d7f7c] pseries_lpar_resize_hpt+0x19c/0x2c0
[c0000000feb03500] [c0000000000788d0] resize_hpt_for_hotplug+0x70/0xd0
[c0000000feb03570] [c000000000e5d278] arch_add_memory+0x58/0xfc
[c0000000feb03610] [c0000000003553a8] devm_memremap_pages+0x5e8/0x8f0
[c0000000feb036c0] [c0000000009c2394] pmem_attach_disk+0x764/0x830
[c0000000feb037d0] [c0000000009a7c38] nvdimm_bus_probe+0x118/0x240
[c0000000feb03860] [c000000000968500] really_probe+0x230/0x4b0
[c0000000feb038f0] [c000000000968aec] driver_probe_device+0x16c/0x1e0
[c0000000feb03970] [c000000000968ca8] __driver_attach+0x148/0x1b0
[c0000000feb039f0] [c0000000009650b0] bus_for_each_dev+0x90/0x130
[c0000000feb03a50] [c000000000967dd4] driver_attach+0x34/0x50
[c0000000feb03a70] [c000000000967068] bus_add_driver+0x1a8/0x360
[c0000000feb03b00] [c00000000096a498] driver_register+0x108/0x170
[c0000000feb03b70] [c0000000009a7400] __nd_driver_register+0xd0/0xf0
[c0000000feb03bd0] [c00000000128aa90] nd_pmem_driver_init+0x34/0x48
[c0000000feb03bf0] [c000000000010a10] do_one_initcall+0x1e0/0x45c
[c0000000feb03cd0] [c00000000122462c] kernel_init_freeable+0x540/0x64c
[c0000000feb03db0] [c00000000001110c] kernel_init+0x2c/0x160
[c0000000feb03e20] [c00000000000bed4] ret_from_kernel_thread+0x5c/0x68
Fix this issue by
1) Requiring all the calls to pseries_lpar_resize_hpt() be made
with cpu_hotplug_lock held.
2) In pseries_lpar_resize_hpt() invoke stop_machine_cpuslocked()
as a consequence of 1)
3) To satisfy 1), in hpt_order_set(), call mmu_hash_ops.resize_hpt()
with cpu_hotplug_lock held.
Reported-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com>
---
v2 -> v3 : Updated the comment for pseries_lpar_resize_hpt()
Updated the commit-log with the full backtrace.
v1 -> v2 : Rebased against powerpc/next instead of linux/master
arch/powerpc/mm/book3s64/hash_utils.c | 9 ++++++++-
arch/powerpc/platforms/pseries/lpar.c | 8 ++++++--
2 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/arch/powerpc/mm/book3s64/hash_utils.c b/arch/powerpc/mm/book3s64/hash_utils.c
index 919a861..d07fcafd 100644
--- a/arch/powerpc/mm/book3s64/hash_utils.c
+++ b/arch/powerpc/mm/book3s64/hash_utils.c
@@ -38,6 +38,7 @@
#include <linux/libfdt.h>
#include <linux/pkeys.h>
#include <linux/hugetlb.h>
+#include <linux/cpu.h>
#include <asm/debugfs.h>
#include <asm/processor.h>
@@ -1928,10 +1929,16 @@ static int hpt_order_get(void *data, u64 *val)
static int hpt_order_set(void *data, u64 val)
{
+ int ret;
+
if (!mmu_hash_ops.resize_hpt)
return -ENODEV;
- return mmu_hash_ops.resize_hpt(val);
+ cpus_read_lock();
+ ret = mmu_hash_ops.resize_hpt(val);
+ cpus_read_unlock();
+
+ return ret;
}
DEFINE_DEBUGFS_ATTRIBUTE(fops_hpt_order, hpt_order_get, hpt_order_set, "%llu\n");
diff --git a/arch/powerpc/platforms/pseries/lpar.c b/arch/powerpc/platforms/pseries/lpar.c
index 1034ef1..557d592 100644
--- a/arch/powerpc/platforms/pseries/lpar.c
+++ b/arch/powerpc/platforms/pseries/lpar.c
@@ -859,7 +859,10 @@ static int pseries_lpar_resize_hpt_commit(void *data)
return 0;
}
-/* Must be called in user context */
+/*
+ * Must be called in process context. The caller must hold the
+ * cpus_lock.
+ */
static int pseries_lpar_resize_hpt(unsigned long shift)
{
struct hpt_resize_state state = {
@@ -913,7 +916,8 @@ static int pseries_lpar_resize_hpt(unsigned long shift)
t1 = ktime_get();
- rc = stop_machine(pseries_lpar_resize_hpt_commit, &state, NULL);
+ rc = stop_machine_cpuslocked(pseries_lpar_resize_hpt_commit,
+ &state, NULL);
t2 = ktime_get();
--
1.9.4
^ permalink raw reply related
* Re: [PATCH] vsprintf: Do not break early boot with probing addresses
From: Petr Mladek @ 2019-05-15 7:53 UTC (permalink / raw)
To: Geert Uytterhoeven
Cc: linux-arch@vger.kernel.org, Sergey Senozhatsky, Heiko Carstens,
linux-s390@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
Rasmus Villemoes, linux-kernel@vger.kernel.org, Steven Rostedt,
Michal Hocko, Sergey Senozhatsky, David Laight, Stephen Rothwell,
Andy Shevchenko, Linus Torvalds, Martin Schwidefsky,
Tobin C . Harding
In-Reply-To: <CAMuHMdUFEwxwQUWg0HNUiz75hP6S7TVGSx7fWXnsQ_3qsVfotQ@mail.gmail.com>
On Wed 2019-05-15 09:23:05, Geert Uytterhoeven wrote:
> Hi Steve,
>
> On Tue, May 14, 2019 at 9:35 PM Steven Rostedt <rostedt@goodmis.org> wrote:
> > On Tue, 14 May 2019 21:13:06 +0200
> > Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> > > > > Do we care about the value? "(-E%u)"?
> > > >
> > > > That too could be confusing. What would (-E22) be considered by a user
> > > > doing an sprintf() on some string. I know that would confuse me, or I
> > > > would think that it was what the %pX displayed, and wonder why it
> > > > displayed it that way. Whereas "(fault)" is quite obvious for any %p
> > > > use case.
> > >
> > > I would immediately understand there's a missing IS_ERR() check in a
> > > function that can return -EINVAL, without having to add a new printk()
> > > to find out what kind of bogus value has been received, and without
> > > having to reboot, and trying to reproduce...
> >
> > I have to ask. Has there actually been a case that you used a %pX and
> > it faulted, and you had to go back to find what the value of the
> > failure was?
>
> If it faulted, the bad pointer value is obvious from the backtrace.
> If the code avoids the fault by verifying the pointer and returning
> "(efault)" instead, the bad pointer value is lost.
>
> Or am I missing something?
Should buggy printk() crash the system?
Another problem is that vsprintf() is called in printk() under
lockbuf_lock. The messages are stored into printk_safe per CPU
buffers. It allows to see the nested messages. But there is still
a bigger risk of missing them than with a "normal" fault.
Finally, various variants of these checks were already used
in "random" printf formats. The only change is that we are
using them consistently everywhere[*] a pointer is accessed.
[*] Just the top level pointer is checked. Some pointer modifiers
are accessing ptr->ptr->val. The lower level pointers are not
checked to avoid too much complexity.
Best Regards,
Petr
^ permalink raw reply
* Re: [PATCH 2/3] arm64: dts: ls1028a: Add PCIe controller DT nodes
From: Arnd Bergmann @ 2019-05-15 8:05 UTC (permalink / raw)
To: Xiaowei Bao
Cc: Mark Rutland, Roy Zang, Lorenzo Pieralisi, DTML, gregkh,
Kate Stewart, linuxppc-dev, linux-pci, Linux Kernel Mailing List,
Kishon, M.h. Lian, Rob Herring, Linux ARM, Philippe Ombredanne,
Bjorn Helgaas, Leo Li, Shawn Guo, Shawn Lin, Mingkai Hu
In-Reply-To: <20190515072747.39941-2-xiaowei.bao@nxp.com>
On Wed, May 15, 2019 at 9:36 AM Xiaowei Bao <xiaowei.bao@nxp.com> wrote:
> Signed-off-by: Xiaowei Bao <xiaowei.bao@nxp.com>
> ---
> arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi | 52 ++++++++++++++++++++++++
> 1 files changed, 52 insertions(+), 0 deletions(-)
>
> diff --git a/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi b/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
> index b045812..50b579b 100644
> --- a/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
> +++ b/arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
> @@ -398,6 +398,58 @@
> status = "disabled";
> };
>
> + pcie@3400000 {
> + compatible = "fsl,ls1028a-pcie";
> + reg = <0x00 0x03400000 0x0 0x00100000 /* controller registers */
> + 0x80 0x00000000 0x0 0x00002000>; /* configuration space */
> + reg-names = "regs", "config";
> + interrupts = <GIC_SPI 108 IRQ_TYPE_LEVEL_HIGH>, /* PME interrupt */
> + <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>; /* aer interrupt */
> + interrupt-names = "pme", "aer";
> + #address-cells = <3>;
> + #size-cells = <2>;
> + device_type = "pci";
> + dma-coherent;
> + num-lanes = <4>;
> + bus-range = <0x0 0xff>;
> + ranges = <0x81000000 0x0 0x00000000 0x80 0x00010000 0x0 0x00010000 /* downstream I/O */
> + 0x82000000 0x0 0x40000000 0x80 0x40000000 0x0 0x40000000>; /* non-prefetchable memory */
Are you sure there is no support for 64-bit BARs or prefetchable memory?
Is this a hardware bug, or something that can be fixed in firmware?
Arnd
^ permalink raw reply
* Re: [PATCH stable 4.9] powerpc/lib: fix book3s/32 boot failure due to code patching
From: Greg KH @ 2019-05-15 8:29 UTC (permalink / raw)
To: Christophe Leroy
Cc: erhard_f, Michael Neuling, linux-kernel, stable@vger.kernel.org,
Paul Mackerras, linuxppc-dev
In-Reply-To: <629c2acb1fcd09c2d2e3352370c3d9853372cf39.1557902321.git.christophe.leroy@c-s.fr>
On Wed, May 15, 2019 at 06:40:47AM +0000, Christophe Leroy wrote:
> [Backport of upstream commit b45ba4a51cde29b2939365ef0c07ad34c8321789]
>
> On powerpc32, patch_instruction() is called by apply_feature_fixups()
> which is called from early_init()
>
> There is the following note in front of early_init():
> * Note that the kernel may be running at an address which is different
> * from the address that it was linked at, so we must use RELOC/PTRRELOC
> * to access static data (including strings). -- paulus
>
> Therefore init_mem_is_free must be accessed with PTRRELOC()
>
> Fixes: 1c38a84d4586 ("powerpc: Avoid code patching freed init sections")
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=203597
> Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
>
> ---
> Can't apply the upstream commit as such due to several other unrelated stuff
> like for instance STRICT_KERNEL_RWX which are missing.
> So instead, using same approach as for commit 252eb55816a6f69ef9464cad303cdb3326cdc61d
Now queued up, thanks.
greg k-h
^ permalink raw reply
* RE: [PATCH] vsprintf: Do not break early boot with probing addresses
From: David Laight @ 2019-05-15 9:00 UTC (permalink / raw)
To: 'Petr Mladek', Steven Rostedt
Cc: linux-arch@vger.kernel.org, Sergey Senozhatsky, Heiko Carstens,
linux-s390@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
Rasmus Villemoes, linux-kernel@vger.kernel.org, Michal Hocko,
Sergey Senozhatsky, Geert Uytterhoeven, Stephen Rothwell,
Andy Shevchenko, Linus Torvalds, Martin Schwidefsky,
Tobin C . Harding
In-Reply-To: <20190515073542.y6ru2nfagtcrpdl7@pathway.suse.cz>
From: Petr Mladek
> Sent: 15 May 2019 08:36
> On Tue 2019-05-14 14:37:51, Steven Rostedt wrote:
> >
> > [ Purple is a nice shade on the bike shed. ;-) ]
> >
> > On Tue, 14 May 2019 11:02:17 +0200
> > Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> >
> > > On Tue, May 14, 2019 at 10:29 AM David Laight <David.Laight@aculab.com> wrote:
> > > > > And I like Steven's "(fault)" idea.
> > > > > How about this:
> > > > >
> > > > > if ptr < PAGE_SIZE -> "(null)"
> > > > > if IS_ERR_VALUE(ptr) -> "(fault)"
> > > > >
> > > > > -ss
> > > >
> > > > Or:
> > > > if (ptr < PAGE_SIZE)
> > > > return ptr ? "(null+)" : "(null)";
> >
> > Hmm, that is useful.
> >
> > > > if IS_ERR_VALUE(ptr)
> > > > return "(errno)"
> >
> > I still prefer "(fault)" as is pretty much all I would expect from a
> > pointer dereference, even if it is just bad parsing of, say, a parsing
> > an MAC address. "fault" is generic enough. "errno" will be confusing,
> > because that's normally a variable not a output.
> >
> > >
> > > Do we care about the value? "(-E%u)"?
> >
> > That too could be confusing. What would (-E22) be considered by a user
> > doing an sprintf() on some string. I know that would confuse me, or I
> > would think that it was what the %pX displayed, and wonder why it
> > displayed it that way. Whereas "(fault)" is quite obvious for any %p
> > use case.
>
> This discussion clearly shows that it is hard to make anyone happy.
>
> I considered switching to "(fault)" because there seems to be more
> people in favor of this.
>
> But there is used also "(einval)" when an unsupported pointer
> modifier is passed. The idea is to show error codes that people
> are familiar with.
>
> It might have been better to use the uppercase "(EFAULT)" and
> "(EINVAL)" to make it more obvious. But I wanted to follow
> the existing style with the lowercase "(null)".
Printing 'fault' when the code was (trying to) validate the
address was ok.
When the only check is for an -errno value it seems wrong as
most invalid addresses will actually fault (and panic).
The reason modern printf generate "(null)" is that it is far too
easy for a diagnostic print to fail to test a pointer.
It also makes it easier when 'throwing in' printf while debugging
to add a single trace that will work regardless of whether a
call had succeeded or not.
With the Linux kernel putting errno values into pointers it
seems likely that most invalid pointers in printf will actaully
be error values.
Printing the value will be helpful during debugging - as a
trace can be put after a call and show the parameters and result.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
^ permalink raw reply
* [PATCH] powerpc: Remove double free
From: Tobin C. Harding @ 2019-05-15 9:07 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman
Cc: Greg Kroah-Hartman, linuxppc-dev, Tobin C. Harding, linux-kernel
kfree() after kobject_put(). Who ever wrote this was on crack.
Fixes: 7e8039795a80 ("powerpc/cacheinfo: Fix kobject memleak")
Signed-off-by: Tobin C. Harding <tobin@kernel.org>
---
FTR
git log --pretty=format:"%h%x09%an%x09%ad%x09%s" | grep 7e8039795a80
7e8039795a80 Tobin C. Harding Tue Apr 30 11:09:23 2019 +1000 powerpc/cacheinfo: Fix kobject memleak
arch/powerpc/kernel/cacheinfo.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/arch/powerpc/kernel/cacheinfo.c b/arch/powerpc/kernel/cacheinfo.c
index f2ed3ef4b129..862e2890bd3d 100644
--- a/arch/powerpc/kernel/cacheinfo.c
+++ b/arch/powerpc/kernel/cacheinfo.c
@@ -767,7 +767,6 @@ static void cacheinfo_create_index_dir(struct cache *cache, int index,
cache_dir->kobj, "index%d", index);
if (rc) {
kobject_put(&index_dir->kobj);
- kfree(index_dir);
return;
}
--
2.21.0
^ permalink raw reply related
* Re: [PATCH] powerpc: Remove double free
From: Christophe Leroy @ 2019-05-15 9:26 UTC (permalink / raw)
To: Tobin C. Harding, Benjamin Herrenschmidt, Paul Mackerras,
Michael Ellerman
Cc: Greg Kroah-Hartman, linuxppc-dev, linux-kernel
In-Reply-To: <20190515090750.30647-1-tobin@kernel.org>
kobject_put() released index_dir->kobj
but who will release 'index' ?
Christophe
Le 15/05/2019 à 11:07, Tobin C. Harding a écrit :
> kfree() after kobject_put(). Who ever wrote this was on crack.
>
> Fixes: 7e8039795a80 ("powerpc/cacheinfo: Fix kobject memleak")
> Signed-off-by: Tobin C. Harding <tobin@kernel.org>
> ---
>
> FTR
>
> git log --pretty=format:"%h%x09%an%x09%ad%x09%s" | grep 7e8039795a80
> 7e8039795a80 Tobin C. Harding Tue Apr 30 11:09:23 2019 +1000 powerpc/cacheinfo: Fix kobject memleak
>
> arch/powerpc/kernel/cacheinfo.c | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/arch/powerpc/kernel/cacheinfo.c b/arch/powerpc/kernel/cacheinfo.c
> index f2ed3ef4b129..862e2890bd3d 100644
> --- a/arch/powerpc/kernel/cacheinfo.c
> +++ b/arch/powerpc/kernel/cacheinfo.c
> @@ -767,7 +767,6 @@ static void cacheinfo_create_index_dir(struct cache *cache, int index,
> cache_dir->kobj, "index%d", index);
> if (rc) {
> kobject_put(&index_dir->kobj);
> - kfree(index_dir);
> return;
> }
>
>
^ permalink raw reply
* [PATCH] powerpc/mm: Drop VM_BUG_ON in get_region_id
From: Aneesh Kumar K.V @ 2019-05-15 9:45 UTC (permalink / raw)
To: npiggin, paulus, mpe; +Cc: Aneesh Kumar K.V, linuxppc-dev
We can call get_region_id without validating the ea value. That means
with a wrong ea value we hit the BUG as below.
kernel BUG at arch/powerpc/include/asm/book3s/64/hash.h:129!
Oops: Exception in kernel mode, sig: 5 [#1]
LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=2048 NUMA pSeries
CPU: 0 PID: 3937 Comm: access_tests Not tainted 5.1.0
....
NIP [c00000000007ba20] do_slb_fault+0x70/0x320
LR [c00000000000896c] data_access_slb_common+0x15c/0x1a0
Fix this by removing the VM_BUG_ON. All callers make sure the returned region id
is valid and error out otherwise.
Fixes: 0034d395f89d ("powerpc/mm/hash64: Map all the kernel regions in the same 0xc range")
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
arch/powerpc/include/asm/book3s/64/hash.h | 2 --
1 file changed, 2 deletions(-)
diff --git a/arch/powerpc/include/asm/book3s/64/hash.h b/arch/powerpc/include/asm/book3s/64/hash.h
index 1d1183048cfd..5486087e64ea 100644
--- a/arch/powerpc/include/asm/book3s/64/hash.h
+++ b/arch/powerpc/include/asm/book3s/64/hash.h
@@ -122,11 +122,9 @@ static inline int get_region_id(unsigned long ea)
if (ea < H_KERN_VIRT_START)
return LINEAR_MAP_REGION_ID;
- VM_BUG_ON(id != 0xc);
BUILD_BUG_ON(NON_LINEAR_REGION_ID(H_VMALLOC_START) != 2);
region_id = NON_LINEAR_REGION_ID(ea);
- VM_BUG_ON(region_id > VMEMMAP_REGION_ID);
return region_id;
}
--
2.21.0
^ permalink raw reply related
* [PATCH] powerpc/pseries: Fix xive=off command line
From: Greg Kurz @ 2019-05-15 10:05 UTC (permalink / raw)
To: Michael Ellerman, Paul Mackerras, Benjamin Herrenschmidt
Cc: pavrampu, linuxppc-dev, linux-kernel, stable,
Cédric Le Goater
On POWER9, if the hypervisor supports XIVE exploitation mode, the guest OS
will unconditionally requests for the XIVE interrupt mode even if XIVE was
deactivated with the kernel command line xive=off. Later on, when the spapr
XIVE init code handles xive=off, it disables XIVE and tries to fall back on
the legacy mode XICS.
This discrepency causes a kernel panic because the hypervisor is configured
to provide the XIVE interrupt mode to the guest :
[ 0.008837] kernel BUG at arch/powerpc/sysdev/xics/xics-common.c:135!
[ 0.008877] Oops: Exception in kernel mode, sig: 5 [#1]
[ 0.008908] LE SMP NR_CPUS=1024 NUMA pSeries
[ 0.008939] Modules linked in:
[ 0.008964] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W 5.0.13-200.fc29.ppc64le #1
[ 0.009018] NIP: c000000001029ab8 LR: c000000001029aac CTR: c0000000018e0000
[ 0.009065] REGS: c0000007f96d7900 TRAP: 0700 Tainted: G W (5.0.13-200.fc29.ppc64le)
[ 0.009119] MSR: 8000000002029033 <SF,VEC,EE,ME,IR,DR,RI,LE> CR: 28000222 XER: 20040000
[ 0.009168] CFAR: c0000000001b1e28 IRQMASK: 0
[ 0.009168] GPR00: c000000001029aac c0000007f96d7b90 c0000000015e8600 0000000000000000
[ 0.009168] GPR04: 0000000000000001 0000000000000000 0000000000000061 646f6d61696e0d0a
[ 0.009168] GPR08: 00000007fd8f0000 0000000000000001 c0000000014c44c0 c0000007f96d76cf
[ 0.009168] GPR12: 0000000000000000 c0000000018e0000 0000000000000001 0000000000000000
[ 0.009168] GPR16: 0000000000000000 0000000000000001 c0000007f96d7c08 c0000000016903d0
[ 0.009168] GPR20: c0000007fffe04e8 ffffffffffffffea c000000001620164 c00000000161fe58
[ 0.009168] GPR24: c000000000ea6c88 c0000000011151a8 00000000006000c0 c0000007f96d7c34
[ 0.009168] GPR28: 0000000000000000 c0000000014b286c c000000001115180 c00000000161dc70
[ 0.009558] NIP [c000000001029ab8] xics_smp_probe+0x38/0x98
[ 0.009590] LR [c000000001029aac] xics_smp_probe+0x2c/0x98
[ 0.009622] Call Trace:
[ 0.009639] [c0000007f96d7b90] [c000000001029aac] xics_smp_probe+0x2c/0x98 (unreliable)
[ 0.009687] [c0000007f96d7bb0] [c000000001033404] pSeries_smp_probe+0x40/0xa0
[ 0.009734] [c0000007f96d7bd0] [c0000000010212a4] smp_prepare_cpus+0x62c/0x6ec
[ 0.009782] [c0000007f96d7cf0] [c0000000010141b8] kernel_init_freeable+0x148/0x448
[ 0.009829] [c0000007f96d7db0] [c000000000010ba4] kernel_init+0x2c/0x148
[ 0.009870] [c0000007f96d7e20] [c00000000000bdd4] ret_from_kernel_thread+0x5c/0x68
[ 0.009916] Instruction dump:
[ 0.009940] 7c0802a6 60000000 7c0802a6 38800002 f8010010 f821ffe1 3c62001c e863b9a0
[ 0.009988] 4b1882d1 60000000 7c690034 5529d97e <0b090000> 3d22001c e929b998 3ce2ff8f
Look for xive=off during prom_init and don't ask for XIVE in this case. One
exception though: if the host only supports XIVE, we still want to boot so
we ignore xive=off.
Similarly, have the spapr XIVE init code to looking at the interrupt mode
negociated during CAS, and ignore xive=off if the hypervisor only supports
XIVE.
Fixes: eac1e731b59e ("powerpc/xive: guest exploitation of the XIVE interrupt controller")
Cc: stable@vger.kernel.org # v4.20
Reported-by: Pavithra R. Prakash <pavrampu@in.ibm.com>
Signed-off-by: Greg Kurz <groug@kaod.org>
---
eac1e731b59e is a v4.16 commit actually but this patch only applies
cleanly to v4.20 and newer. If needed I can send a backport for
older versions.
---
arch/powerpc/kernel/prom_init.c | 16 +++++++++++-
arch/powerpc/sysdev/xive/spapr.c | 52 +++++++++++++++++++++++++++++++++++++-
2 files changed, 66 insertions(+), 2 deletions(-)
diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c
index 523bb99d7676..c8f7eb845927 100644
--- a/arch/powerpc/kernel/prom_init.c
+++ b/arch/powerpc/kernel/prom_init.c
@@ -172,6 +172,7 @@ static unsigned long __prombss prom_tce_alloc_end;
#ifdef CONFIG_PPC_PSERIES
static bool __prombss prom_radix_disable;
+static bool __prombss prom_xive_disable;
#endif
struct platform_support {
@@ -808,6 +809,12 @@ static void __init early_cmdline_parse(void)
}
if (prom_radix_disable)
prom_debug("Radix disabled from cmdline\n");
+
+ opt = prom_strstr(prom_cmd_line, "xive=off");
+ if (opt) {
+ prom_xive_disable = true;
+ prom_debug("XIVE disabled from cmdline\n");
+ }
#endif /* CONFIG_PPC_PSERIES */
}
@@ -1216,10 +1223,17 @@ static void __init prom_parse_xive_model(u8 val,
switch (val) {
case OV5_FEAT(OV5_XIVE_EITHER): /* Either Available */
prom_debug("XIVE - either mode supported\n");
- support->xive = true;
+ support->xive = !prom_xive_disable;
break;
case OV5_FEAT(OV5_XIVE_EXPLOIT): /* Only Exploitation mode */
prom_debug("XIVE - exploitation mode supported\n");
+ if (prom_xive_disable) {
+ /*
+ * If we __have__ to do XIVE, we're better off ignoring
+ * the command line rather than not booting.
+ */
+ prom_printf("WARNING: Ignoring cmdline option xive=off\n");
+ }
support->xive = true;
break;
case OV5_FEAT(OV5_XIVE_LEGACY): /* Only Legacy mode */
diff --git a/arch/powerpc/sysdev/xive/spapr.c b/arch/powerpc/sysdev/xive/spapr.c
index 575db3b06a6b..2e2d1b8f810f 100644
--- a/arch/powerpc/sysdev/xive/spapr.c
+++ b/arch/powerpc/sysdev/xive/spapr.c
@@ -20,6 +20,7 @@
#include <linux/cpumask.h>
#include <linux/mm.h>
#include <linux/delay.h>
+#include <linux/libfdt.h>
#include <asm/prom.h>
#include <asm/io.h>
@@ -663,6 +664,55 @@ static bool xive_get_max_prio(u8 *max_prio)
return true;
}
+static const u8 *get_vec5_feature(unsigned int index)
+{
+ unsigned long root, chosen;
+ int size;
+ const u8 *vec5;
+
+ root = of_get_flat_dt_root();
+ chosen = of_get_flat_dt_subnode_by_name(root, "chosen");
+ if (chosen == -FDT_ERR_NOTFOUND)
+ return NULL;
+
+ vec5 = of_get_flat_dt_prop(chosen, "ibm,architecture-vec-5", &size);
+ if (!vec5)
+ return NULL;
+
+ if (size <= index)
+ return NULL;
+
+ return vec5 + index;
+}
+
+static bool xive_spapr_disabled(void)
+{
+ const u8 *vec5_xive;
+
+ vec5_xive = get_vec5_feature(OV5_INDX(OV5_XIVE_SUPPORT));
+ if (vec5_xive) {
+ u8 val;
+
+ val = *vec5_xive & OV5_FEAT(OV5_XIVE_SUPPORT);
+ switch (val) {
+ case OV5_FEAT(OV5_XIVE_EITHER):
+ case OV5_FEAT(OV5_XIVE_LEGACY):
+ break;
+ case OV5_FEAT(OV5_XIVE_EXPLOIT):
+ /* Hypervisor only supports XIVE */
+ if (xive_cmdline_disabled)
+ pr_warn("WARNING: Ignoring cmdline option xive=off\n");
+ return false;
+ default:
+ pr_warn("%s: Unknown xive support option: 0x%x\n",
+ __func__, val);
+ break;
+ }
+ }
+
+ return xive_cmdline_disabled;
+}
+
bool __init xive_spapr_init(void)
{
struct device_node *np;
@@ -675,7 +725,7 @@ bool __init xive_spapr_init(void)
const __be32 *reg;
int i;
- if (xive_cmdline_disabled)
+ if (xive_spapr_disabled())
return false;
pr_devel("%s()\n", __func__);
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox