It may be that the offset of the first program header lies inside the mdt's
filesize, in this case the loader would incorrectly assume that the bins
were not split. The loading would then continue on to fail for split bins.
This change updates the logic used by the mdt loader to understand whether
the firmware images are split or not. It figures this out by checking if
each programs header's segment lies within the file or not.
Signed-off-by: Melody Olvera <[email protected]>
Signed-off-by: Gokul krishna Krishnakumar <[email protected]>
---
V4 addresses the comments from V3.
V3 is separated out from [1] and includes
changes addressing comments from that patch set.
[1] https://lore.kernel.org/all/[email protected]/
---
drivers/soc/qcom/mdt_loader.c | 27 +++++++++++++++++++++++++--
1 file changed, 25 insertions(+), 2 deletions(-)
diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c
index 33dd8c315eb7..9270992728d4 100644
--- a/drivers/soc/qcom/mdt_loader.c
+++ b/drivers/soc/qcom/mdt_loader.c
@@ -258,6 +258,26 @@ int qcom_mdt_pas_init(struct device *dev, const struct firmware *fw,
}
EXPORT_SYMBOL_GPL(qcom_mdt_pas_init);
+static bool qcom_mdt_bins_are_split(const struct firmware *fw, const char* fw_name)
+{
+ const struct elf32_phdr *phdrs;
+ const struct elf32_hdr *ehdr;
+ uint64_t seg_start, seg_end;
+ int i;
+
+ ehdr = (struct elf32_hdr *)fw->data;
+ phdrs = (struct elf32_phdr *)(ehdr + 1);
+
+ for (i = 0; i < ehdr->e_phnum; i++) {
+ seg_start = phdrs[i].p_offset;
+ seg_end = phdrs[i].p_offset + phdrs[i].p_filesz;
+ if (seg_start > fw->size || seg_end > fw->size)
+ return true;
+ }
+
+ return false;
+}
+
static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
const char *fw_name, int pas_id, void *mem_region,
phys_addr_t mem_phys, size_t mem_size,
@@ -270,6 +290,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
phys_addr_t min_addr = PHYS_ADDR_MAX;
ssize_t offset;
bool relocate = false;
+ bool is_split;
void *ptr;
int ret = 0;
int i;
@@ -277,6 +298,8 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
if (!fw || !mem_region || !mem_phys || !mem_size)
return -EINVAL;
+
+ is_split = qcom_mdt_bins_are_split(fw, fw_name);
ehdr = (struct elf32_hdr *)fw->data;
phdrs = (struct elf32_phdr *)(ehdr + 1);
@@ -330,8 +353,8 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
ptr = mem_region + offset;
- if (phdr->p_filesz && phdr->p_offset < fw->size &&
- phdr->p_offset + phdr->p_filesz <= fw->size) {
+
+ if (phdr->p_filesz && !is_split) {
/* Firmware is large enough to be non-split */
if (phdr->p_offset + phdr->p_filesz > fw->size) {
dev_err(dev, "file %s segment %d would be truncated\n",
--
2.39.2
On Tue, Apr 04, 2023 at 10:54:17AM -0700, Gokul krishna Krishnakumar wrote:
> It may be that the offset of the first program header lies inside the mdt's
> filesize, in this case the loader would incorrectly assume that the bins
> were not split. The loading would then continue on to fail for split bins.
What does "continue on to fail for split bins" actually mean? In what
way does it fail?
> This change updates the logic used by the mdt loader to understand whether
> the firmware images are split or not. It figures this out by checking if
> each programs header's segment lies within the file or not.
>
> Signed-off-by: Melody Olvera <[email protected]>
> Signed-off-by: Gokul krishna Krishnakumar <[email protected]>
The first S-o-b should be that of the author (unless Co-developed-by)
and in [1] you're the author and Melody provide her S-o-b to certify
the origin of her contribution.
> ---
> V4 addresses the comments from V3.
That's nice, but not very helpful. Please spell out what you changed.
>
> V3 is separated out from [1] and includes
> changes addressing comments from that patch set.
>
> [1] https://lore.kernel.org/all/[email protected]/
> ---
> drivers/soc/qcom/mdt_loader.c | 27 +++++++++++++++++++++++++--
> 1 file changed, 25 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c
> index 33dd8c315eb7..9270992728d4 100644
> --- a/drivers/soc/qcom/mdt_loader.c
> +++ b/drivers/soc/qcom/mdt_loader.c
> @@ -258,6 +258,26 @@ int qcom_mdt_pas_init(struct device *dev, const struct firmware *fw,
> }
> EXPORT_SYMBOL_GPL(qcom_mdt_pas_init);
>
> +static bool qcom_mdt_bins_are_split(const struct firmware *fw, const char* fw_name)
> +{
> + const struct elf32_phdr *phdrs;
> + const struct elf32_hdr *ehdr;
> + uint64_t seg_start, seg_end;
> + int i;
> +
> + ehdr = (struct elf32_hdr *)fw->data;
> + phdrs = (struct elf32_phdr *)(ehdr + 1);
> +
> + for (i = 0; i < ehdr->e_phnum; i++) {
> + seg_start = phdrs[i].p_offset;
> + seg_end = phdrs[i].p_offset + phdrs[i].p_filesz;
> + if (seg_start > fw->size || seg_end > fw->size)
> + return true;
> + }
> +
> + return false;
> +}
> +
> static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
> const char *fw_name, int pas_id, void *mem_region,
> phys_addr_t mem_phys, size_t mem_size,
> @@ -270,6 +290,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
> phys_addr_t min_addr = PHYS_ADDR_MAX;
> ssize_t offset;
> bool relocate = false;
> + bool is_split;
> void *ptr;
> int ret = 0;
> int i;
> @@ -277,6 +298,8 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
> if (!fw || !mem_region || !mem_phys || !mem_size)
> return -EINVAL;
>
> +
Double empty lines here?
> + is_split = qcom_mdt_bins_are_split(fw, fw_name);
> ehdr = (struct elf32_hdr *)fw->data;
> phdrs = (struct elf32_phdr *)(ehdr + 1);
>
> @@ -330,8 +353,8 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
>
> ptr = mem_region + offset;
>
> - if (phdr->p_filesz && phdr->p_offset < fw->size &&
> - phdr->p_offset + phdr->p_filesz <= fw->size) {
> +
> + if (phdr->p_filesz && !is_split) {
This looks much better now, thanks.
Regards,
Bjorn
> /* Firmware is large enough to be non-split */
> if (phdr->p_offset + phdr->p_filesz > fw->size) {
> dev_err(dev, "file %s segment %d would be truncated\n",
> --
> 2.39.2
>
On 4/4/2023 12:05 PM, Bjorn Andersson wrote:
> On Tue, Apr 04, 2023 at 10:54:17AM -0700, Gokul krishna Krishnakumar wrote:
>> It may be that the offset of the first program header lies inside the mdt's
>> filesize, in this case the loader would incorrectly assume that the bins
>> were not split. The loading would then continue on to fail for split bins.
>
> What does "continue on to fail for split bins" actually mean? In what
> way does it fail?
> Authentication fails in the firmware for the split bins because the
current conditional check do not go through the entire mdt file for a
given firmware.
Thanks,
Gokul
>> This change updates the logic used by the mdt loader to understand whether
>> the firmware images are split or not. It figures this out by checking if
>> each programs header's segment lies within the file or not.
>>
>> Signed-off-by: Melody Olvera <[email protected]>
>> Signed-off-by: Gokul krishna Krishnakumar <[email protected]>
>
> The first S-o-b should be that of the author (unless Co-developed-by)
> and in [1] you're the author and Melody provide her S-o-b to certify
> the origin of her contribution.
>
>> ---
>> V4 addresses the comments from V3.
>
> That's nice, but not very helpful. Please spell out what you changed.
>
>>
>> V3 is separated out from [1] and includes
>> changes addressing comments from that patch set.
>>
>> [1] https://lore.kernel.org/all/[email protected]/
>> ---
>> drivers/soc/qcom/mdt_loader.c | 27 +++++++++++++++++++++++++--
>> 1 file changed, 25 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c
>> index 33dd8c315eb7..9270992728d4 100644
>> --- a/drivers/soc/qcom/mdt_loader.c
>> +++ b/drivers/soc/qcom/mdt_loader.c
>> @@ -258,6 +258,26 @@ int qcom_mdt_pas_init(struct device *dev, const struct firmware *fw,
>> }
>> EXPORT_SYMBOL_GPL(qcom_mdt_pas_init);
>>
>> +static bool qcom_mdt_bins_are_split(const struct firmware *fw, const char* fw_name)
>> +{
>> + const struct elf32_phdr *phdrs;
>> + const struct elf32_hdr *ehdr;
>> + uint64_t seg_start, seg_end;
>> + int i;
>> +
>> + ehdr = (struct elf32_hdr *)fw->data;
>> + phdrs = (struct elf32_phdr *)(ehdr + 1);
>> +
>> + for (i = 0; i < ehdr->e_phnum; i++) {
>> + seg_start = phdrs[i].p_offset;
>> + seg_end = phdrs[i].p_offset + phdrs[i].p_filesz;
>> + if (seg_start > fw->size || seg_end > fw->size)
>> + return true;
>> + }
>> +
>> + return false;
>> +}
>> +
>> static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
>> const char *fw_name, int pas_id, void *mem_region,
>> phys_addr_t mem_phys, size_t mem_size,
>> @@ -270,6 +290,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
>> phys_addr_t min_addr = PHYS_ADDR_MAX;
>> ssize_t offset;
>> bool relocate = false;
>> + bool is_split;
>> void *ptr;
>> int ret = 0;
>> int i;
>> @@ -277,6 +298,8 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
>> if (!fw || !mem_region || !mem_phys || !mem_size)
>> return -EINVAL;
>>
>> +
>
> Double empty lines here?
>
>> + is_split = qcom_mdt_bins_are_split(fw, fw_name);
>> ehdr = (struct elf32_hdr *)fw->data;
>> phdrs = (struct elf32_phdr *)(ehdr + 1);
>>
>> @@ -330,8 +353,8 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw,
>>
>> ptr = mem_region + offset;
>>
>> - if (phdr->p_filesz && phdr->p_offset < fw->size &&
>> - phdr->p_offset + phdr->p_filesz <= fw->size) {
>> +
>> + if (phdr->p_filesz && !is_split) {
>
> This looks much better now, thanks.
>
> Regards,
> Bjorn
>
>> /* Firmware is large enough to be non-split */
>> if (phdr->p_offset + phdr->p_filesz > fw->size) {
>> dev_err(dev, "file %s segment %d would be truncated\n",
>> --
>> 2.39.2
>>