Support getting the SONAME from program headers.

rust-minidump · Apr 30, 2024 · 8938433 · 8938433
1 parent f0b4480
commit 8938433
Show file tree

Hide file tree

Showing 2 changed files with 210 additions and 59 deletions.
diff --git a/src/linux/errors.rs b/src/linux/errors.rs
@@ -266,9 +266,11 @@ pub enum ModuleReaderError {
     NoStrTab,
     #[error("no build id note sections")]
     NoSectionNote,
-    #[error("the ELF file contains no sections")]
+    #[error("the ELF data contains no program headers")]
+    NoProgramHeaders,
+    #[error("the ELF data contains no sections")]
     NoSections,
-    #[error("the ELF file does not have a .text section from which to generate a build id")]
+    #[error("the ELF data does not have a .text section from which to generate a build id")]
     NoTextSection,
     #[error(
         "failed to calculate build id\n\
@@ -289,4 +291,13 @@ pub enum ModuleReaderError {
     NoSoNameEntry,
     #[error("no dynamic linking information section")]
     NoDynamicSection,
+    #[error(
+        "failed to retrieve soname\n\
+    ... from program headers: {program_headers}\n\
+    ... from sections: {section}"
+    )]
+    NoSoName {
+        program_headers: Box<Self>,
+        section: Box<Self>,
+    },
 }
diff --git a/src/linux/module_reader.rs b/src/linux/module_reader.rs
@@ -11,7 +11,14 @@ const NOTE_SECTION_NAME: &[u8] = b".note.gnu.build-id\0";
 pub trait ModuleMemory {
     type Memory: std::ops::Deref<Target = [u8]>;
 
+    /// Read memory from the module.
     fn read_module_memory(&self, offset: u64, length: u64) -> std::io::Result<Self::Memory>;
+
+    /// Whether the module memory is from a module loaded in the address space of a program.
+    /// The default implementation assumes this to be true.
+    fn is_loaded_in_program(&self) -> bool {
+        true
+    }
 }
 
 impl<'a> ModuleMemory for &'a [u8] {
@@ -28,6 +35,22 @@ impl<'a> ModuleMemory for &'a [u8] {
     }
 }
 
+/// Indicate that a ModuleMemory implementation is read directly from a file (the module is not
+/// loaded by the dynamic linker).
+pub struct ModuleMemoryFromFile<T>(T);
+
+impl<T: ModuleMemory> ModuleMemory for ModuleMemoryFromFile<T> {
+    type Memory = T::Memory;
+
+    fn read_module_memory(&self, offset: u64, length: u64) -> std::io::Result<Self::Memory> {
+        self.0.read_module_memory(offset, length)
+    }
+
+    fn is_loaded_in_program(&self) -> bool {
+        false
+    }
+}
+
 fn read<T: ModuleMemory>(mem: &T, offset: u64, length: u64) -> Result<T::Memory, Error> {
     mem.read_module_memory(offset, length)
         .map_err(|error| Error::ReadModuleMemory {
@@ -73,7 +96,7 @@ fn section_header_with_name<'a>(
     for header in section_headers {
         let sh_name = header.sh_name as u64;
         if sh_name >= strtab_section_header.sh_size {
-            log::warn!("invalid sh_name offset");
+            log::warn!("invalid sh_name offset for {:?}", name);
             continue;
         }
         if sh_name + name.len() as u64 >= strtab_section_header.sh_size {
@@ -124,15 +147,58 @@ impl ReadFromModule for BuildId {
     }
 }
 
+struct DynIter<'a> {
+    data: &'a [u8],
+    offset: usize,
+    ctx: Ctx,
+}
+
+impl<'a> DynIter<'a> {
+    pub fn new(data: &'a [u8], ctx: Ctx) -> Self {
+        DynIter {
+            data,
+            offset: 0,
+            ctx,
+        }
+    }
+}
+
+impl<'a> Iterator for DynIter<'a> {
+    type Item = Result<elf::dynamic::Dyn, Error>;
+
+    fn next(&mut self) -> Option<Self::Item> {
+        use scroll::Pread;
+        let dyn_: elf::dynamic::Dyn = match self.data.gread_with(&mut self.offset, self.ctx) {
+            Ok(v) => v,
+            Err(e) => return Some(Err(e.into())),
+        };
+        if dyn_.d_tag == elf::dynamic::DT_NULL {
+            None
+        } else {
+            Some(Ok(dyn_))
+        }
+    }
+}
+
 /// The module SONAME.
 #[derive(Default, Clone, Debug)]
 pub struct SoName(pub String);
 
 impl ReadFromModule for SoName {
     fn read_from_module(module_memory: impl ModuleMemory) -> Result<Self, Error> {
-        ModuleReader::new(module_memory)
-            .and_then(|r| r.soname())
-            .map(SoName)
+        let reader = ModuleReader::new(module_memory)?;
+        let program_headers = match reader.soname_from_program_headers() {
+            Ok(v) => return Ok(SoName(v)),
+            Err(e) => Box::new(e),
+        };
+        let section = match reader.soname_from_sections() {
+            Ok(v) => return Ok(SoName(v)),
+            Err(e) => Box::new(e),
+        };
+        Err(Error::NoSoName {
+            program_headers,
+            section,
+        })
     }
 }
 
@@ -158,7 +224,41 @@ impl<T: ModuleMemory> ModuleReader<T> {
         })
     }
 
-    pub fn soname(&self) -> Result<String, Error> {
+    /// Read the SONAME using program headers to locate dynamic library information.
+    pub fn soname_from_program_headers(&self) -> Result<String, Error> {
+        let program_headers = self.read_program_headers()?;
+
+        let dynamic_segment_header = program_headers
+            .iter()
+            .find(|h| h.p_type == elf::program_header::PT_DYNAMIC)
+            .ok_or(Error::NoDynamicSection)?;
+
+        let dynamic_section: &[u8] = &self.read_segment(dynamic_segment_header)?;
+
+        let mut soname_strtab_offset = None;
+        let mut strtab_addr = None;
+        let mut strtab_size = None;
+        for dyn_ in DynIter::new(dynamic_section, self.context) {
+            let dyn_ = dyn_?;
+            match dyn_.d_tag {
+                elf::dynamic::DT_SONAME => soname_strtab_offset = Some(dyn_.d_val),
+                elf::dynamic::DT_STRTAB => strtab_addr = Some(dyn_.d_val),
+                elf::dynamic::DT_STRSZ => strtab_size = Some(dyn_.d_val),
+                _ => (),
+            }
+        }
+
+        match (strtab_addr, strtab_size, soname_strtab_offset) {
+            (None, _, _) | (_, None, _) => Err(Error::NoDynStrSection),
+            (_, _, None) => Err(Error::NoSoNameEntry),
+            (Some(addr), Some(size), Some(offset)) => {
+                self.read_name_from_strtab(addr, size, offset)
+            }
+        }
+    }
+
+    /// Read the SONAME using section headers to locate dynamic library information.
+    pub fn soname_from_sections(&self) -> Result<String, Error> {
         let section_headers = self.read_section_headers()?;
 
         let dynamic_section_header = section_headers
@@ -180,50 +280,30 @@ impl<T: ModuleMemory> ModuleReader<T> {
 
         let dynamic_section: &[u8] = &read(
             &self.module_memory,
-            dynamic_section_header.sh_offset,
+            self.section_offset(&dynamic_section_header),
             dynamic_section_header.sh_size,
         )?;
 
-        let mut offset = 0;
-        loop {
-            use scroll::Pread;
-            let dyn_: elf::dynamic::Dyn = dynamic_section.gread_with(&mut offset, self.context)?;
+        for dyn_ in DynIter::new(dynamic_section, self.context) {
+            let dyn_ = dyn_?;
             if dyn_.d_tag == elf::dynamic::DT_SONAME {
-                let strtab_offset = dyn_.d_val;
-                if strtab_offset < dynstr_section_header.sh_size {
-                    let name = read(
-                        &self.module_memory,
-                        dynstr_section_header.sh_offset + strtab_offset,
-                        dynstr_section_header.sh_size - strtab_offset,
-                    )?;
-                    return CStr::from_bytes_until_nul(&name)
-                        .map(|s| s.to_string_lossy().into_owned())
-                        .map_err(|_| Error::StrTabNoNulByte);
+                let name_offset = dyn_.d_val;
+                if name_offset < dynstr_section_header.sh_size {
+                    return self.read_name_from_strtab(
+                        self.section_offset(&dynstr_section_header),
+                        dynstr_section_header.sh_size,
+                        name_offset,
+                    );
                 }
             }
-            if dyn_.d_tag == elf::dynamic::DT_NULL {
-                break;
-            }
         }
+
         Err(Error::NoSoNameEntry)
     }
 
     /// Read the build id from a program header note.
     pub fn build_id_from_program_headers(&self) -> Result<Vec<u8>, Error> {
-        if self.header.e_phoff == 0 {
-            return Err(Error::NoProgramHeaderNote);
-        }
-        let program_headers_data = read(
-            &self.module_memory,
-            self.header.e_phoff,
-            self.header.e_phentsize as u64 * self.header.e_phnum as u64,
-        )?;
-        let program_headers = elf::ProgramHeader::parse(
-            &program_headers_data,
-            0,
-            self.header.e_phnum as usize,
-            self.context,
-        )?;
+        let program_headers = self.read_program_headers()?;
         for header in program_headers {
             if header.p_type != elf::program_header::PT_NOTE {
                 continue;
@@ -272,6 +352,58 @@ impl<T: ModuleMemory> ModuleReader<T> {
         Ok(build_id_from_bytes(&text_data))
     }
 
+    fn read_segment(&self, header: &elf::ProgramHeader) -> Result<T::Memory, Error> {
+        let (offset, size) = if self.module_memory.is_loaded_in_program() {
+            (header.p_vaddr, header.p_memsz)
+        } else {
+            (header.p_offset, header.p_filesz)
+        };
+
+        read(&self.module_memory, offset, size)
+    }
+
+    fn read_name_from_strtab(
+        &self,
+        strtab_offset: u64,
+        strtab_size: u64,
+        name_offset: u64,
+    ) -> Result<String, Error> {
+        let name = read(
+            &self.module_memory,
+            strtab_offset + name_offset,
+            strtab_size - name_offset,
+        )?;
+        return CStr::from_bytes_until_nul(&name)
+            .map(|s| s.to_string_lossy().into_owned())
+            .map_err(|_| Error::StrTabNoNulByte);
+    }
+
+    fn section_offset(&self, header: &elf::SectionHeader) -> u64 {
+        if self.module_memory.is_loaded_in_program() {
+            header.sh_addr
+        } else {
+            header.sh_offset
+        }
+    }
+
+    fn read_program_headers(&self) -> Result<elf::ProgramHeaders, Error> {
+        if self.header.e_phoff == 0 {
+            return Err(Error::NoProgramHeaders);
+        }
+        let program_headers_data = read(
+            &self.module_memory,
+            self.header.e_phoff,
+            self.header.e_phentsize as u64 * self.header.e_phnum as u64,
+        )?;
+        let program_headers = elf::ProgramHeader::parse(
+            &program_headers_data,
+            0,
+            self.header.e_phnum as usize,
+            self.context,
+        )?;
+        Ok(program_headers)
+    }
+
     fn read_section_headers(&self) -> Result<elf::SectionHeaders, Error> {
         if self.header.e_shoff == 0 {
             return Err(Error::NoSections);
@@ -336,32 +468,32 @@ mod test {
     /// * section header: .dynstr
     /// * note header (build id note)
     /// * shstrtab
-    /// * dynamic (SONAME)
+    /// * dynamic (SONAME/STRTAB/STRSZ)
     /// * dynstr (SONAME string = libfoo.so.1)
     /// * program (calls exit(0))
-    const TINY_ELF: &[u8] = &[
+    const TINY_ELF: ModuleMemoryFromFile<&[u8]> = ModuleMemoryFromFile(&[
         0x7f, 0x45, 0x4c, 0x46, 0x02, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-        0x00, 0x02, 0x00, 0x3e, 0x00, 0x01, 0x00, 0x00, 0x00, 0xea, 0x02, 0x40, 0x00, 0x00, 0x00,
+        0x00, 0x02, 0x00, 0x3e, 0x00, 0x01, 0x00, 0x00, 0x00, 0x0a, 0x03, 0x40, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xe8, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x40, 0x00, 0x38, 0x00, 0x03, 0x00, 0x40, 0x00,
-        0x06, 0x00, 0x03, 0x00, 0x01, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0xea, 0x02, 0x00,
-        0x00, 0x00, 0x00, 0x00, 0x00, 0xea, 0x02, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+        0x06, 0x00, 0x03, 0x00, 0x01, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x0a, 0x03, 0x00,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x03, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x07,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x68, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x68, 0x02, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0xbd, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xbd, 0x02, 0x40,
-        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x20, 0x00,
-        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x40, 0x00,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
-        0x01, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xea, 0x02, 0x40,
-        0x00, 0x00, 0x00, 0x00, 0x00, 0xea, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x07, 0x00,
+        0x01, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x03, 0x40,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x03, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x07, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x10, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x07, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
@@ -374,11 +506,11 @@ mod test {
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00,
         0x00, 0x03, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xbd, 0x02, 0x40, 0x00, 0x00, 0x00,
-        0x00, 0x00, 0xbd, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x00,
+        0x00, 0x00, 0xbd, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x2d, 0x00, 0x00,
-        0x00, 0x03, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xdd, 0x02,
-        0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0xdd, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d,
+        0x00, 0x03, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xfd, 0x02,
+        0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0xfd, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
         0x00, 0x04, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x47, 0x4e,
@@ -387,12 +519,13 @@ mod test {
         0x2e, 0x67, 0x6e, 0x75, 0x2e, 0x62, 0x75, 0x69, 0x6c, 0x64, 0x2d, 0x69, 0x64, 0x00, 0x2e,
         0x73, 0x68, 0x73, 0x74, 0x72, 0x74, 0x61, 0x62, 0x00, 0x2e, 0x64, 0x79, 0x6e, 0x61, 0x6d,
         0x69, 0x63, 0x00, 0x2e, 0x64, 0x79, 0x6e, 0x73, 0x74, 0x72, 0x00, 0x0e, 0x00, 0x00, 0x00,
-        0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x6c,
-        0x69, 0x62, 0x66, 0x6f, 0x6f, 0x2e, 0x73, 0x6f, 0x2e, 0x31, 0x00, 0x6a, 0x3c, 0x58, 0x31,
-        0xff, 0x0f, 0x05, 0x66, 0x6f, 0x6f, 0x2e, 0x73, 0x6f, 0x2e, 0x31, 0x00, 0x6a, 0x3c, 0x58,
-        0x31, 0xff, 0x0f, 0x05, 0x05,
-    ];
+        0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0xfd, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x00,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+        0x00, 0x6c, 0x69, 0x62, 0x66, 0x6f, 0x6f, 0x2e, 0x73, 0x6f, 0x2e, 0x31, 0x00, 0x6a, 0x3c,
+        0x58, 0x31, 0xff, 0x0f, 0x05,
+    ]);
 
     #[test]
     fn build_id_program_headers() {
@@ -425,9 +558,16 @@ mod test {
     }
 
     #[test]
-    fn soname() {
+    fn soname_program_headers() {
+        let reader = ModuleReader::new(TINY_ELF).unwrap();
+        let soname = reader.soname_from_program_headers().unwrap();
+        assert_eq!(soname, "libfoo.so.1");
+    }
+
+    #[test]
+    fn soname_section() {
         let reader = ModuleReader::new(TINY_ELF).unwrap();
-        let soname = reader.soname().unwrap();
+        let soname = reader.soname_from_sections().unwrap();
         assert_eq!(soname, "libfoo.so.1");
     }
 }