From: Arnaud POULIQUEN <arnaud.pouliquen@st.com>
To: "Clément Leger" <cleger@kalray.eu>
Cc: Ohad Ben-Cohen <ohad@wizery.com>,
Bjorn Andersson <bjorn.andersson@linaro.org>,
Jonathan Corbet <corbet@lwn.net>, Shawn Guo <shawnguo@kernel.org>,
Sascha Hauer <s.hauer@pengutronix.de>,
linux-remoteproc <linux-remoteproc@vger.kernel.org>,
Pengutronix Kernel Team <kernel@pengutronix.de>,
Fabio Estevam <festevam@gmail.com>,
NXP Linux Team <linux-imx@nxp.com>,
Andy Gross <agross@kernel.org>,
Patrice Chotard <patrice.chotard@st.com>,
linux-doc <linux-doc@vger.kernel.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
linux-arm-kernel <linux-arm-kernel@lists.infradead.org>,
linux-arm-msm <linux-arm-msm@vger.kernel.org>,
Loic PALLARDY <loic.pallardy@st.com>, s-anna <s-anna@ti.com>,
Mathieu Poirier <mathieu.poirier@linaro.org>
Subject: Re: [PATCH v4 1/5] remoteproc: Use u64 len for da_to_va
Date: Tue, 11 Feb 2020 18:24:07 +0100 [thread overview]
Message-ID: <63c2949f-c1d6-c0aa-b123-906630e3f028@st.com> (raw)
In-Reply-To: <884697376.3644142.1581439161953.JavaMail.zimbra@kalray.eu>
On 2/11/20 5:39 PM, Clément Leger wrote:
> Hi Arnaud,
>
> ----- On 11 Feb, 2020, at 16:53, Arnaud Pouliquen arnaud.pouliquen@st.com wrote:
>
>> On 2/10/20 5:22 PM, Clement Leger wrote:
>>> With upcoming changes in elf loader for elf64 support, section size will
>>> be a u64. When used with da_to_va, this will potentially lead to
>>> overflow if using the current "int" type for len argument. Change
>>> da_to_va prototype to use a u64 for len and fix all users of this
>>> function.
>>>
>>> Signed-off-by: Clement Leger <cleger@kalray.eu>
>>> ---
>>> drivers/remoteproc/imx_rproc.c | 11 ++++++-----
>>> drivers/remoteproc/keystone_remoteproc.c | 4 ++--
>>> drivers/remoteproc/qcom_q6v5_adsp.c | 2 +-
>>> drivers/remoteproc/qcom_q6v5_mss.c | 2 +-
>>> drivers/remoteproc/qcom_q6v5_pas.c | 2 +-
>>> drivers/remoteproc/qcom_q6v5_wcss.c | 2 +-
>>> drivers/remoteproc/qcom_wcnss.c | 2 +-
>>> drivers/remoteproc/remoteproc_core.c | 2 +-
>>> drivers/remoteproc/remoteproc_internal.h | 2 +-
>>> drivers/remoteproc/st_slim_rproc.c | 4 ++--
>>> drivers/remoteproc/wkup_m3_rproc.c | 4 ++--
>>> include/linux/remoteproc.h | 2 +-
>>> 12 files changed, 20 insertions(+), 19 deletions(-)
>>>
>>> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
>>> index 3e72b6f38d4b..f497f5b49b18 100644
>>> --- a/drivers/remoteproc/imx_rproc.c
>>> +++ b/drivers/remoteproc/imx_rproc.c
>>> @@ -186,7 +186,7 @@ static int imx_rproc_stop(struct rproc *rproc)
>>> }
>>>
>>> static int imx_rproc_da_to_sys(struct imx_rproc *priv, u64 da,
>>> - int len, u64 *sys)
>>> + u64 len, u64 *sys)
>>> {
>>> const struct imx_rproc_dcfg *dcfg = priv->dcfg;
>>> int i;
>>> @@ -203,19 +203,19 @@ static int imx_rproc_da_to_sys(struct imx_rproc *priv, u64
>>> da,
>>> }
>>> }
>>>
>>> - dev_warn(priv->dev, "Translation failed: da = 0x%llx len = 0x%x\n",
>>> + dev_warn(priv->dev, "Translation failed: da = 0x%llx len = 0x%llx\n",
>>> da, len);
>>> return -ENOENT;
>>> }
>>>
>>> -static void *imx_rproc_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *imx_rproc_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct imx_rproc *priv = rproc->priv;
>>> void *va = NULL;
>>> u64 sys;
>>> int i;
>>>
>>> - if (len <= 0)
>>> + if (len == 0)
>>> return NULL;
>>>
>>> /*
>>> @@ -235,7 +235,8 @@ static void *imx_rproc_da_to_va(struct rproc *rproc, u64 da,
>>> int len)
>>> }
>>> }
>>>
>>> - dev_dbg(&rproc->dev, "da = 0x%llx len = 0x%x va = 0x%p\n", da, len, va);
>>> + dev_dbg(&rproc->dev, "da = 0x%llx len = 0x%llx va = 0x%p\n",
>>> + da, len, va);
>>>
>>> return va;
>>> }
>>> diff --git a/drivers/remoteproc/keystone_remoteproc.c
>>> b/drivers/remoteproc/keystone_remoteproc.c
>>> index 5c4658f00b3d..466093f48814 100644
>>> --- a/drivers/remoteproc/keystone_remoteproc.c
>>> +++ b/drivers/remoteproc/keystone_remoteproc.c
>>> @@ -246,7 +246,7 @@ static void keystone_rproc_kick(struct rproc *rproc, int
>>> vqid)
>>> * can be used either by the remoteproc core for loading (when using kernel
>>> * remoteproc loader), or by any rpmsg bus drivers.
>>> */
>>> -static void *keystone_rproc_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *keystone_rproc_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct keystone_rproc *ksproc = rproc->priv;
>>> void __iomem *va = NULL;
>>> @@ -255,7 +255,7 @@ static void *keystone_rproc_da_to_va(struct rproc *rproc,
>>> u64 da, int len)
>>> size_t size;
>>> int i;
>>>
>>> - if (len <= 0)
>>> + if (len == 0)
>>> return NULL;
>>>
>>> for (i = 0; i < ksproc->num_mems; i++) {
>>> diff --git a/drivers/remoteproc/qcom_q6v5_adsp.c
>>> b/drivers/remoteproc/qcom_q6v5_adsp.c
>>> index e953886b2eb7..7518e67a49e5 100644
>>> --- a/drivers/remoteproc/qcom_q6v5_adsp.c
>>> +++ b/drivers/remoteproc/qcom_q6v5_adsp.c
>>> @@ -270,7 +270,7 @@ static int adsp_stop(struct rproc *rproc)
>>> return ret;
>>> }
>>>
>>> -static void *adsp_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *adsp_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct qcom_adsp *adsp = (struct qcom_adsp *)rproc->priv;
>>> int offset;
>>> diff --git a/drivers/remoteproc/qcom_q6v5_mss.c
>>> b/drivers/remoteproc/qcom_q6v5_mss.c
>>> index 471128a2e723..248febde6fc1 100644
>>> --- a/drivers/remoteproc/qcom_q6v5_mss.c
>>> +++ b/drivers/remoteproc/qcom_q6v5_mss.c
>>> @@ -1148,7 +1148,7 @@ static int q6v5_stop(struct rproc *rproc)
>>> return 0;
>>> }
>>>
>>> -static void *q6v5_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *q6v5_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct q6v5 *qproc = rproc->priv;
>>> int offset;
>>> diff --git a/drivers/remoteproc/qcom_q6v5_pas.c
>>> b/drivers/remoteproc/qcom_q6v5_pas.c
>>> index db4b3c4bacd7..cf2cd609c90d 100644
>>> --- a/drivers/remoteproc/qcom_q6v5_pas.c
>>> +++ b/drivers/remoteproc/qcom_q6v5_pas.c
>>> @@ -159,7 +159,7 @@ static int adsp_stop(struct rproc *rproc)
>>> return ret;
>>> }
>>>
>>> -static void *adsp_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *adsp_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct qcom_adsp *adsp = (struct qcom_adsp *)rproc->priv;
>>> int offset;
>>> diff --git a/drivers/remoteproc/qcom_q6v5_wcss.c
>>> b/drivers/remoteproc/qcom_q6v5_wcss.c
>>> index f93e1e4a1cc0..3a6b82a16961 100644
>>> --- a/drivers/remoteproc/qcom_q6v5_wcss.c
>>> +++ b/drivers/remoteproc/qcom_q6v5_wcss.c
>>> @@ -406,7 +406,7 @@ static int q6v5_wcss_stop(struct rproc *rproc)
>>> return 0;
>>> }
>>>
>>> -static void *q6v5_wcss_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *q6v5_wcss_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct q6v5_wcss *wcss = rproc->priv;
>>> int offset;
>>> diff --git a/drivers/remoteproc/qcom_wcnss.c b/drivers/remoteproc/qcom_wcnss.c
>>> index dc135754bb9c..f893219e45a8 100644
>>> --- a/drivers/remoteproc/qcom_wcnss.c
>>> +++ b/drivers/remoteproc/qcom_wcnss.c
>>> @@ -287,7 +287,7 @@ static int wcnss_stop(struct rproc *rproc)
>>> return ret;
>>> }
>>>
>>> -static void *wcnss_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *wcnss_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct qcom_wcnss *wcnss = (struct qcom_wcnss *)rproc->priv;
>>> int offset;
>>> diff --git a/drivers/remoteproc/remoteproc_core.c
>>> b/drivers/remoteproc/remoteproc_core.c
>>> index 307df98347ba..9e6d3c6a60ee 100644
>>> --- a/drivers/remoteproc/remoteproc_core.c
>>> +++ b/drivers/remoteproc/remoteproc_core.c
>>> @@ -185,7 +185,7 @@ EXPORT_SYMBOL(rproc_va_to_pa);
>>> * here the output of the DMA API for the carveouts, which should be more
>>> * correct.
>>> */
>>> -void *rproc_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +void *rproc_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>
>> This function is exported, don't see any update in consequence...
>> references:
>> https://elixir.bootlin.com/linux/v5.6-rc1/ident/rproc_da_to_va
>> For instance the function rproc_trace_read use it. it quite strange that my gcc
>> does not warns for the cast but i suppose that some could.
>
> Agreed, even if len should never have been a signed type since it can't be
> negative. I will try to fix all callers.
>
>> An indirect consequence is that the len field in rproc_mem_entry struct should
>> probably been updated to u64 to be aligned.
>
> Ok, I will do that once we settle on the type of len.
>
>>
>> I'm still wondering about the use of size_t instead,which seems more rational
>> from my window.
>> So i you or Mathieu remember it was decided to use u64, please could remind me
>> the arguments?
>
> I tried to find the notes of a meeting we had for OpenAMP but I did not found
> them. Anyway, the argument was coming from Tomas or someone else, (I can't
> remember) talking about a 32 bits CPU executing code on a 64 bits accelerator.
> In that case, the size_t type could fail due to being only 32bits on the host
> CPU but larger than 4G.
>
> However, I can't say if it's a real usecase or not... All I can say is
> that keeping it open is probably better if one day somebody comes with such
> architecture.
>
>> As an alternative a check should be added for 32 bits processors to ensure that
>> the size is not higher than
>> its address range capability...
>
> Agreed.
> I was even thinking about a mecanism for remoteproc drivers to declare the type
> of supported elfs files (such as EM_*, ELFCLASS* and other needed thing).
> Or should it be supported by overriding .sanity_check in drivers to reject
> elf64 for instance ?
>
> Since elf is a "specific format" and that rproc can support other formats,
> I did not want to add a specific elf_sanity_check field to rproc ops.
Agree, platform driver as to check the format it can support, elf, bin, properitay format....
To check the size, the da_to_va ops seems a good candidate as called for loaded segments and memories.
So ok to delegate the check to the platform driver.
Regards
Arnaud
>
> Regards,
>
> Clément
>
>>
>> Regards
>> Arnaud
>>
>>> {
>>> struct rproc_mem_entry *carveout;
>>> void *ptr = NULL;
>>> diff --git a/drivers/remoteproc/remoteproc_internal.h
>>> b/drivers/remoteproc/remoteproc_internal.h
>>> index 493ef9262411..004867061721 100644
>>> --- a/drivers/remoteproc/remoteproc_internal.h
>>> +++ b/drivers/remoteproc/remoteproc_internal.h
>>> @@ -50,7 +50,7 @@ void rproc_exit_sysfs(void);
>>> void rproc_free_vring(struct rproc_vring *rvring);
>>> int rproc_alloc_vring(struct rproc_vdev *rvdev, int i);
>>>
>>> -void *rproc_da_to_va(struct rproc *rproc, u64 da, int len);
>>> +void *rproc_da_to_va(struct rproc *rproc, u64 da, u64 len);
>>> phys_addr_t rproc_va_to_pa(void *cpu_addr);
>>> int rproc_trigger_recovery(struct rproc *rproc);
>>>
>>> diff --git a/drivers/remoteproc/st_slim_rproc.c
>>> b/drivers/remoteproc/st_slim_rproc.c
>>> index 04492fead3c8..fc01cd879b60 100644
>>> --- a/drivers/remoteproc/st_slim_rproc.c
>>> +++ b/drivers/remoteproc/st_slim_rproc.c
>>> @@ -174,7 +174,7 @@ static int slim_rproc_stop(struct rproc *rproc)
>>> return 0;
>>> }
>>>
>>> -static void *slim_rproc_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *slim_rproc_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct st_slim_rproc *slim_rproc = rproc->priv;
>>> void *va = NULL;
>>> @@ -191,7 +191,7 @@ static void *slim_rproc_da_to_va(struct rproc *rproc, u64
>>> da, int len)
>>> }
>>> }
>>>
>>> - dev_dbg(&rproc->dev, "da = 0x%llx len = 0x%x va = 0x%pK\n",
>>> + dev_dbg(&rproc->dev, "da = 0x%llx len = 0x%llx va = 0x%pK\n",
>>> da, len, va);
>>>
>>> return va;
>>> diff --git a/drivers/remoteproc/wkup_m3_rproc.c
>>> b/drivers/remoteproc/wkup_m3_rproc.c
>>> index 3984e585c847..91485b467407 100644
>>> --- a/drivers/remoteproc/wkup_m3_rproc.c
>>> +++ b/drivers/remoteproc/wkup_m3_rproc.c
>>> @@ -80,14 +80,14 @@ static int wkup_m3_rproc_stop(struct rproc *rproc)
>>> return 0;
>>> }
>>>
>>> -static void *wkup_m3_rproc_da_to_va(struct rproc *rproc, u64 da, int len)
>>> +static void *wkup_m3_rproc_da_to_va(struct rproc *rproc, u64 da, u64 len)
>>> {
>>> struct wkup_m3_rproc *wkupm3 = rproc->priv;
>>> void *va = NULL;
>>> int i;
>>> u32 offset;
>>>
>>> - if (len <= 0)
>>> + if (len == 0)
>>> return NULL;
>>>
>>> for (i = 0; i < WKUPM3_MEM_MAX; i++) {
>>> diff --git a/include/linux/remoteproc.h b/include/linux/remoteproc.h
>>> index 16ad66683ad0..f84bd5fe0211 100644
>>> --- a/include/linux/remoteproc.h
>>> +++ b/include/linux/remoteproc.h
>>> @@ -374,7 +374,7 @@ struct rproc_ops {
>>> int (*start)(struct rproc *rproc);
>>> int (*stop)(struct rproc *rproc);
>>> void (*kick)(struct rproc *rproc, int vqid);
>>> - void * (*da_to_va)(struct rproc *rproc, u64 da, int len);
>>> + void * (*da_to_va)(struct rproc *rproc, u64 da, u64 len);
>>> int (*parse_fw)(struct rproc *rproc, const struct firmware *fw);
>>> int (*handle_rsc)(struct rproc *rproc, u32 rsc_type, void *rsc,
>>> int offset, int avail);
next prev parent reply other threads:[~2020-02-11 17:24 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20200129163013.GA16538@xps15>
2020-02-04 14:33 ` [PATCH v2 1/2] remoteproc: Use u64 len for da_to_va Clement Leger
2020-02-04 14:33 ` [PATCH v2 2/2] remoteproc: Add elf64 support in elf loader Clement Leger
2020-02-04 16:27 ` [PATCH v2 1/2] remoteproc: Use u64 len for da_to_va Arnaud POULIQUEN
2020-02-04 17:10 ` Clément Leger
2020-02-04 17:42 ` Arnaud POULIQUEN
2020-02-04 17:44 ` [PATCH v3 0/2] remoteproc: Add elf64 support to elf loader Clement Leger
2020-02-04 17:44 ` [PATCH v3 1/2] remoteproc: Use u64 len for da_to_va Clement Leger
2020-02-05 21:06 ` Mathieu Poirier
2020-02-04 17:44 ` [PATCH v3 2/2] remoteproc: Add elf64 support in elf loader Clement Leger
2020-02-05 22:49 ` Mathieu Poirier
2020-02-06 8:37 ` Clément Leger
2020-02-06 15:05 ` Clément Leger
2020-02-06 17:48 ` Mathieu Poirier
2020-02-07 7:57 ` Clément Leger
2020-02-10 16:22 ` [PATCH v4 0/5] remoteproc: Add elf64 support Clement Leger
2020-02-10 16:22 ` [PATCH v4 1/5] remoteproc: Use u64 len for da_to_va Clement Leger
2020-02-11 15:53 ` Arnaud POULIQUEN
2020-02-11 16:39 ` Clément Leger
2020-02-11 17:24 ` Arnaud POULIQUEN [this message]
2020-02-11 22:37 ` Mathieu Poirier
2020-02-12 10:37 ` Clément Leger
2020-02-12 21:59 ` Mathieu Poirier
2020-02-18 10:10 ` Clément Leger
2020-02-18 17:01 ` Mathieu Poirier
2020-02-10 16:22 ` [PATCH v4 2/5] remoteproc: Use u64 type for boot_addr Clement Leger
2020-02-10 16:22 ` [PATCH v4 3/5] remoteproc: Add elf helpers to access elf64 and elf32 fields Clement Leger
2020-02-10 16:22 ` [PATCH v4 4/5] remoteproc: Add elf64 support in elf loader Clement Leger
2020-02-10 16:22 ` [PATCH v4 5/5] remoteproc: Adapt coredump to generate correct elf type Clement Leger
2020-02-11 23:11 ` Mathieu Poirier
2020-02-11 15:57 ` [PATCH v4 0/5] remoteproc: Add elf64 support Arnaud POULIQUEN
2020-02-11 23:12 ` Mathieu Poirier
2020-02-12 8:15 ` Arnaud POULIQUEN
2020-03-02 9:38 ` [PATCH v5 0/8] " Clement Leger
2020-03-02 9:38 ` [PATCH v5 1/8] remoteproc: Use size_t type for len in da_to_va Clement Leger
2020-03-02 23:06 ` Bjorn Andersson
2020-03-09 17:52 ` Mathieu Poirier
2020-03-27 7:37 ` Oleksij Rempel
2020-03-02 9:38 ` [PATCH v5 2/8] remoteproc: Use size_t instead of int for rproc_mem_entry len Clement Leger
2020-03-02 23:07 ` Bjorn Andersson
2020-03-09 19:21 ` Mathieu Poirier
2020-03-02 9:38 ` [PATCH v5 3/8] remoteproc: Use u64 type for boot_addr Clement Leger
2020-03-02 23:08 ` Bjorn Andersson
2020-03-09 19:52 ` Mathieu Poirier
2020-03-10 7:59 ` Clément Leger
2020-03-10 19:32 ` Bjorn Andersson
2020-03-02 9:38 ` [PATCH v5 4/8] remoteproc: Add elf helpers to access elf64 and elf32 fields Clement Leger
2020-03-02 23:12 ` Bjorn Andersson
2020-03-09 19:56 ` Mathieu Poirier
2020-03-02 9:38 ` [PATCH v5 5/8] remoteproc: Rename rproc_elf_sanity_check for elf32 Clement Leger
2020-03-02 23:13 ` Bjorn Andersson
2020-03-03 8:02 ` Clément Leger
2020-03-10 0:00 ` Bjorn Andersson
2020-03-10 15:20 ` Mathieu Poirier
2020-03-10 15:38 ` Clément Leger
2020-03-10 19:18 ` Mathieu Poirier
2020-03-10 19:30 ` Bjorn Andersson
2020-03-02 9:39 ` [PATCH v5 6/8] remoteproc: Add elf64 support in elf loader Clement Leger
2020-03-02 9:39 ` [PATCH v5 7/8] remoteproc: Allow overriding only sanity_check Clement Leger
2020-03-02 9:39 ` [PATCH v5 8/8] remoteproc: Adapt coredump to generate correct elf type Clement Leger
2020-03-03 22:01 ` Bjorn Andersson
2020-03-09 20:32 ` Mathieu Poirier
2020-03-09 23:57 ` Bjorn Andersson
2020-03-10 8:12 ` Clément Leger
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=63c2949f-c1d6-c0aa-b123-906630e3f028@st.com \
--to=arnaud.pouliquen@st.com \
--cc=agross@kernel.org \
--cc=bjorn.andersson@linaro.org \
--cc=cleger@kalray.eu \
--cc=corbet@lwn.net \
--cc=festevam@gmail.com \
--cc=kernel@pengutronix.de \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-imx@nxp.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-remoteproc@vger.kernel.org \
--cc=loic.pallardy@st.com \
--cc=mathieu.poirier@linaro.org \
--cc=ohad@wizery.com \
--cc=patrice.chotard@st.com \
--cc=s-anna@ti.com \
--cc=s.hauer@pengutronix.de \
--cc=shawnguo@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).